Fix azure ai vector store (#2396)

Co-authored-by: Deshraj Yadav <deshrajdry@gmail.com>

Makefile

@@ -12,7 +12,7 @@ install:
 
 install_all:
 	poetry install
-	poetry run pip install groq together boto3 litellm ollama chromadb sentence_transformers vertexai \
+	poetry run pip install groq together boto3 litellm ollama chromadb weaviate weaviate-client sentence_transformers vertexai \
 	                        google-generativeai elasticsearch opensearch-py vecs
 
 # Format code with ruff

docs/components/vectordbs/dbs/azure_ai_search.mdx

@@ -1,12 +1,14 @@
-[Azure AI Search](https://learn.microsoft.com/en-us/azure/search/search-what-is-azure-search/) (formerly known as "Azure Cognitive Search") provides secure information retrieval at scale over user-owned content in traditional and generative AI search applications.
+# Azure AI Search
 
-### Usage
+[Azure AI Search](https://learn.microsoft.com/azure/search/search-what-is-azure-search/) (formerly known as "Azure Cognitive Search") provides secure information retrieval at scale over user-owned content in traditional and generative AI search applications.
+
+## Usage
 
 ```python
 import os
 from mem0 import Memory
 
-os.environ["OPENAI_API_KEY"] = "sk-xx"   #this key is used for embedding purpose
+os.environ["OPENAI_API_KEY"] = "sk-xx"   # This key is used for embedding purpose
 
 config = {
     "vector_store": {
@@ -15,8 +17,8 @@ config = {
             "service_name": "ai-search-test",
             "api_key": "*****",
             "collection_name": "mem0", 
-            "embedding_model_dims": 1536 ,
-            "use_compression": False
+            "embedding_model_dims": 1536,
+            "compression_type": "none"
         }
     }
 }
@@ -25,20 +27,46 @@ m = Memory.from_config(config)
 messages = [
     {"role": "user", "content": "I'm planning to watch a movie tonight. Any recommendations?"},
     {"role": "assistant", "content": "How about a thriller movies? They can be quite engaging."},
-    {"role": "user", "content": "I’m not a big fan of thriller movies but I love sci-fi movies."},
+    {"role": "user", "content": "I'm not a big fan of thriller movies but I love sci-fi movies."},
     {"role": "assistant", "content": "Got it! I'll avoid thriller recommendations and suggest sci-fi movies in the future."}
 ]
 m.add(messages, user_id="alice", metadata={"category": "movies"})
 ```
 
-### Config
+# Using binary compression for large vector collections
+config = {
+    "vector_store": {
+        "provider": "azure_ai_search",
+        "config": {
+            "service_name": "ai-search-test",
+            "api_key": "*****",
+            "collection_name": "mem0", 
+            "embedding_model_dims": 1536,
+            "compression_type": "binary",
+            "use_float16": True  # Use half precision for storage efficiency
+        }
+    }
+}
+```
 
-Let's see the available parameters for the `qdrant` config:
-service_name (str): Azure Cognitive Search service name.  
-| Parameter | Description | Default Value |
-| --- | --- | --- |
-| `service_name` | Azure AI Search service name | `None` |
-| `api_key` | API key of the Azure AI Search service | `None` |
-| `collection_name` | The name of the collection/index to store the vectors, it will be created automatically if not exist | `mem0` |
-| `embedding_model_dims` | Dimensions of the embedding model | `1536` |
-| `use_compression` | Use scalar quantization vector compression | False |
+## Configuration Parameters
+
+| Parameter | Description | Default Value | Options |
+| --- | --- | --- | --- |
+| `service_name` | Azure AI Search service name | Required | - |
+| `api_key` | API key of the Azure AI Search service | Required | - |
+| `collection_name` | The name of the collection/index to store vectors | `mem0` | Any valid index name |
+| `embedding_model_dims` | Dimensions of the embedding model | `1536` | Any integer value |
+| `compression_type` | Type of vector compression to use | `none` | `none`, `scalar`, `binary` |
+| `use_float16` | Store vectors in half precision (Edm.Half) | `False` | `True`, `False` |
+
+## Notes on Configuration Options
+
+- **compression_type**: 
+  - `none`: No compression, uses full vector precision
+  - `scalar`: Scalar quantization with reasonable balance of speed and accuracy
+  - `binary`: Binary quantization for maximum compression with some accuracy trade-off
+
+- **use_float16**: Using half precision (float16) reduces storage requirements but may slightly impact accuracy. Useful for very large vector collections.
+
+- **Filterable Fields**: The implementation automatically extracts `user_id`, `run_id`, and `agent_id` fields from payloads for filtering.

docs/components/vectordbs/dbs/elasticsearch.mdx

@@ -54,6 +54,7 @@ Let's see the available parameters for the `elasticsearch` config:
 | `password`             | Password for basic authentication                  | `None`        |
 | `verify_certs`         | Whether to verify SSL certificates                 | `True`        |
 | `auto_create_index`    | Whether to automatically create the index          | `True`        |
+| `custom_search_query`  | Function returning a custom search query            | `None`        |
 
 ### Features
 
@@ -62,3 +63,46 @@ Let's see the available parameters for the `elasticsearch` config:
 - Multiple authentication methods (Basic Auth, API Key)
 - Automatic index creation with optimized mappings for vector search
 - Memory isolation through payload filtering
+- Custom search query function to customize the search query
+
+### Custom Search Query
+
+The `custom_search_query` parameter allows you to customize the search query when `Memory.search` is called.  
+  
+__Example__  
+```python
+import os
+from typing import List, Optional, Dict
+from mem0 import Memory
+
+def custom_search_query(query: List[float], limit: int, filters: Optional[Dict]) -> Dict:
+    return {
+        "knn": {
+            "field": "vector", 
+            "query_vector": query, 
+            "k": limit, 
+            "num_candidates": limit * 2
+        }
+    }
+
+os.environ["OPENAI_API_KEY"] = "sk-xx"
+
+config = {
+    "vector_store": {
+        "provider": "elasticsearch",
+        "config": {
+            "collection_name": "mem0",
+            "host": "localhost",
+            "port": 9200,
+            "embedding_model_dims": 1536,
+            "custom_search_query": custom_search_query
+        }
+    }
+}
+```
+It should be a function that takes the following parameters:
+- `query`: a query vector used in `Memory.search`
+- `limit`: a number of results used in `Memory.search`
+- `filters`: a dictionary of key-value pairs used in `Memory.search`. You can add custom pairs for the custom search query.  
+  
+The function should return a query body for the Elasticsearch search API.

docs/components/vectordbs/dbs/weaviate.mdx

@@ -0,0 +1,47 @@
+[Weaviate](https://weaviate.io/) is an open-source vector search engine. It allows efficient storage and retrieval of high-dimensional vector embeddings, enabling powerful search and retrieval capabilities.
+
+
+### Installation
+```bash
+pip install weaviate weaviate-client
+```
+
+### Usage
+
+```python Python
+import os
+from mem0 import Memory
+
+os.environ["OPENAI_API_KEY"] = "sk-xx"
+
+config = {
+    "vector_store": {
+        "provider": "weaviate",
+        "config": {
+            "collection_name": "test",
+            "cluster_url": "http://localhost:8080",
+            "auth_client_secret": None,
+        }
+    }
+}
+
+m = Memory.from_config(config)
+messages = [
+    {"role": "user", "content": "I'm planning to watch a movie tonight. Any recommendations?"},
+    {"role": "assistant", "content": "How about a thriller movie? They can be quite engaging."},
+    {"role": "user", "content": "I’m not a big fan of thriller movies but I love sci-fi movies."},
+    {"role": "assistant", "content": "Got it! I'll avoid thriller recommendations and suggest sci-fi movies in the future."}
+]
+m.add(messages, user_id="alice", metadata={"category": "movies"})
+```
+
+### Config
+
+Let's see the available parameters for the `weaviate` config:
+
+| Parameter | Description | Default Value |
+| --- | --- | --- |
+| `collection_name` | The name of the collection to store the vectors | `mem0` |
+| `embedding_model_dims` | Dimensions of the embedding model | `1536` |
+| `cluster_url` | URL for the Weaviate server | `None` |
+| `auth_client_secret` | API key for Weaviate authentication | `None` |

docs/components/vectordbs/overview.mdx

@@ -25,6 +25,7 @@ See the list of supported vector databases below.
   <Card title="OpenSearch" href="/components/vectordbs/dbs/opensearch"></Card>
   <Card title="Supabase" href="/components/vectordbs/dbs/supabase"></Card>
   <Card title="Vertex AI Vector Search" href="/components/vectordbs/dbs/vertex_ai_vector_search"></Card>
+  <Card title="Weaviate" href="/components/vectordbs/dbs/weaviate"></Card>
 </CardGroup>
 
 ## Usage

docs/docs.json

@@ -47,6 +47,7 @@
                     "pages": [
                       "features/platform-overview",
                       "features/advanced-retrieval",
+                      "features/contextual-add",
                       "features/multimodal-support",
                       "features/selective-memory",
                       "features/custom-categories",
@@ -71,7 +72,8 @@
                     "icon": "wrench",
                     "pages": [
                       "features/openai_compatibility",
-                      "features/custom-prompts",
+                      "features/custom-fact-extraction-prompt",
+                      "features/custom-update-memory-prompt",
                       "open-source/multimodal-support",
                       "open-source/features/rest-api"
                     ]
@@ -130,7 +132,8 @@
                           "components/vectordbs/dbs/elasticsearch",
                           "components/vectordbs/dbs/opensearch",
                           "components/vectordbs/dbs/supabase",
-                          "components/vectordbs/dbs/vertex_ai_vector_search"
+                          "components/vectordbs/dbs/vertex_ai_vector_search",
+                          "components/vectordbs/dbs/weaviate"
                         ]
                       }
                     ]
@@ -185,7 +188,9 @@
                   "examples/chrome-extension",
                   "examples/document-writing",
                   "examples/multimodal-demo",
-                  "examples/personalized-deep-research"
+                  "examples/personalized-deep-research",
+                  "examples/mem0-agentic-tool",
+                  "examples/openai-inbuilt-tools"
                 ]
               }
             ]
@@ -333,6 +338,9 @@
     "posthog": {
       "apiKey": "phc_hgJkUVJFYtmaJqrvf6CYN67TIQ8yhXAkWzUn9AMU4yX",
       "apiHost": "https://mango.mem0.ai"
+    },
+    "intercom": {
+      "appId": "jjv2r0tt"
     }
   }
 }

docs/examples/mem0-agentic-tool.mdx

@@ -0,0 +1,226 @@
+---
+title: Mem0 as an Agentic Tool
+---
+
+Integrate Mem0's memory capabilities with OpenAI's Agents SDK to create AI agents with persistent memory.
+You can create agents that remember past conversations and use that context to provide better responses.
+
+## Installation
+
+First, install the required packages:
+```bash
+pip install mem0ai pydantic openai-agents
+```
+
+You'll also need a custom agents framework for this implementation.
+
+## Setting Up Environment Variables
+
+Store your Mem0 API key as an environment variable:
+
+```bash
+export MEM0_API_KEY="your_mem0_api_key"
+```
+
+Or in your Python script:
+
+```python
+import os
+os.environ["MEM0_API_KEY"] = "your_mem0_api_key"
+```
+
+## Code Structure
+
+The integration consists of three main components:
+
+1. **Context Manager**: Defines user context for memory operations
+2. **Memory Tools**: Functions to add, search, and retrieve memories
+3. **Memory Agent**: An agent configured to use these memory tools
+
+## Step-by-Step Implementation
+
+### 1. Import Dependencies
+
+```python
+from __future__ import annotations
+import os
+import asyncio
+from pydantic import BaseModel
+try:
+    from mem0 import AsyncMemoryClient
+except ImportError:
+    raise ImportError("mem0 is not installed. Please install it using 'pip install mem0ai'.")
+from agents import (
+    Agent,
+    ItemHelpers,
+    MessageOutputItem,
+    RunContextWrapper,
+    Runner,
+    ToolCallItem,
+    ToolCallOutputItem,
+    TResponseInputItem,
+    function_tool,
+)
+```
+
+### 2. Define Memory Context
+
+```python
+class Mem0Context(BaseModel):
+    user_id: str | None = None
+```
+
+### 3. Initialize the Mem0 Client
+
+```python
+client = AsyncMemoryClient(api_key=os.getenv("MEM0_API_KEY"))
+```
+
+### 4. Create Memory Tools
+
+#### Add to Memory
+
+```python
+@function_tool
+async def add_to_memory(
+    context: RunContextWrapper[Mem0Context],
+    content: str,
+) -> str:
+    """
+    Add a message to Mem0
+    Args:
+        content: The content to store in memory.
+    """
+    messages = [{"role": "user", "content": content}]
+    user_id = context.context.user_id or "default_user"
+    await client.add(messages, user_id=user_id)
+    return f"Stored message: {content}"
+```
+
+#### Search Memory
+
+```python
+@function_tool
+async def search_memory(
+    context: RunContextWrapper[Mem0Context],
+    query: str,
+) -> str:
+    """
+    Search for memories in Mem0
+    Args:
+        query: The search query.
+    """
+    user_id = context.context.user_id or "default_user"
+    memories = await client.search(query, user_id=user_id, output_format="v1.1")
+    results = '\n'.join([result["memory"] for result in memories["results"]])
+    return str(results)
+```
+
+#### Get All Memories
+
+```python
+@function_tool
+async def get_all_memory(
+    context: RunContextWrapper[Mem0Context],
+) -> str:
+    """Retrieve all memories from Mem0"""
+    user_id = context.context.user_id or "default_user"
+    memories = await client.get_all(user_id=user_id, output_format="v1.1")
+    results = '\n'.join([result["memory"] for result in memories["results"]])
+    return str(results)
+```
+
+### 5. Configure the Memory Agent
+
+```python
+memory_agent = Agent[Mem0Context](
+    name="Memory Assistant",
+    instructions="""You are a helpful assistant with memory capabilities. You can:
+    1. Store new information using add_to_memory
+    2. Search existing information using search_memory
+    3. Retrieve all stored information using get_all_memory
+    When users ask questions:
+    - If they want to store information, use add_to_memory
+    - If they're searching for specific information, use search_memory
+    - If they want to see everything stored, use get_all_memory""",
+    tools=[add_to_memory, search_memory, get_all_memory],
+)
+```
+
+### 6. Implement the Main Runtime Loop
+
+```python
+async def main():
+    current_agent: Agent[Mem0Context] = memory_agent
+    input_items: list[TResponseInputItem] = []
+    context = Mem0Context()
+    while True:
+        user_input = input("Enter your message (or 'quit' to exit): ")
+        if user_input.lower() == 'quit':
+            break
+        input_items.append({"content": user_input, "role": "user"})
+        result = await Runner.run(current_agent, input_items, context=context)
+        for new_item in result.new_items:
+            agent_name = new_item.agent.name
+            if isinstance(new_item, MessageOutputItem):
+                print(f"{agent_name}: {ItemHelpers.text_message_output(new_item)}")
+            elif isinstance(new_item, ToolCallItem):
+                print(f"{agent_name}: Calling a tool")
+            elif isinstance(new_item, ToolCallOutputItem):
+                print(f"{agent_name}: Tool call output: {new_item.output}")
+            else:
+                print(f"{agent_name}: Skipping item: {new_item.__class__.__name__}")
+        input_items = result.to_input_list()
+
+if __name__ == "__main__":
+    asyncio.run(main())
+```
+
+## Usage Examples
+
+### Storing Information
+
+```
+User: Remember that my favorite color is blue
+Agent: Calling a tool
+Agent: Tool call output: Stored message: my favorite color is blue
+Agent: I've stored that your favorite color is blue in my memory. I'll remember that for future conversations.
+```
+
+### Searching Memory
+
+```
+User: What's my favorite color?
+Agent: Calling a tool
+Agent: Tool call output: my favorite color is blue
+Agent: Your favorite color is blue, based on what you've told me earlier.
+```
+
+### Retrieving All Memories
+
+```
+User: What do you know about me?
+Agent: Calling a tool
+Agent: Tool call output: favorite color is blue
+my birthday is on March 15
+Agent: Based on our previous conversations, I know that:
+1. Your favorite color is blue
+2. Your birthday is on March 15
+```
+
+## Advanced Configuration
+
+### Custom User IDs
+
+You can specify different user IDs to maintain separate memory stores for multiple users:
+
+```python
+context = Mem0Context(user_id="user123")
+```
+
+
+## Resources
+
+- [Mem0 Documentation](https://docs.mem0.ai)
+- [Mem0 Dashboard](https://app.mem0.ai/dashboard)
+- [API Reference](https://docs.mem0.ai/api-reference)

docs/examples/openai-inbuilt-tools.mdx

@@ -0,0 +1,312 @@
+---
+title: OpenAI Inbuilt Tools
+---
+
+Integrate Mem0’s memory capabilities with OpenAI’s Inbuilt Tools to create AI agents with persistent memory.
+
+## Getting Started
+
+### Installation
+
+```bash
+npm install mem0ai openai zod
+```
+
+## Environment Setup
+
+Save your Mem0 and OpenAI API keys in a `.env` file:
+
+```
+MEM0_API_KEY=your_mem0_api_key
+OPENAI_API_KEY=your_openai_api_key
+```
+
+Get your Mem0 API key from the [Mem0 Dashboard](https://app.mem0.ai/dashboard/api-keys).
+
+### Configuration
+
+```javascript
+const mem0Config = {
+    apiKey: process.env.MEM0_API_KEY,
+    user_id: "sample-user",
+};
+
+const openAIClient = new OpenAI();
+const mem0Client = new MemoryClient(mem0Config);
+```
+
+### Adding Memories
+
+Store user preferences, past interactions, or any relevant information:
+<CodeGroup>
+```javascript JavaScript
+async function addUserPreferences() {
+    const mem0Client = new MemoryClient(mem0Config);
+    
+    const userPreferences = "I Love BMW, Audi and Porsche. I Hate Mercedes. I love Red cars and Maroon cars. I have a budget of 120K to 150K USD. I like Audi the most.";
+    
+    await mem0Client.add([{
+        role: "user",
+        content: userPreferences,
+    }], mem0Config);
+}
+
+await addUserPreferences();
+```
+
+```json Output (Memories)
+ [
+  {
+    "id": "ff9f3367-9e83-415d-b9c5-dc8befd9a4b4",
+    "data": { "memory": "Loves BMW, Audi, and Porsche" },
+    "event": "ADD"
+  },
+  {
+    "id": "04172ce6-3d7b-45a3-b4a1-ee9798593cb4",
+    "data": { "memory": "Hates Mercedes" },
+    "event": "ADD"
+  },
+  {
+    "id": "db363a5d-d258-4953-9e4c-777c120de34d",
+    "data": { "memory": "Loves red cars and maroon cars" },
+    "event": "ADD"
+  },
+  {
+    "id": "5519aaad-a2ac-4c0d-81d7-0d55c6ecdba8",
+    "data": { "memory": "Has a budget of 120K to 150K USD" },
+    "event": "ADD"
+  },
+  {
+    "id": "523b7693-7344-4563-922f-5db08edc8634",
+    "data": { "memory": "Likes Audi the most" },
+    "event": "ADD"
+  }
+]
+```
+</CodeGroup>
+### Retrieving Memories
+
+Search for relevant memories based on the current user input:
+
+```javascript
+const relevantMemories = await mem0Client.search(userInput, mem0Config);
+```
+
+### Structured Responses with Zod
+
+Define structured response schemas to get consistent output formats:
+
+```javascript
+// Define the schema for a car recommendation
+const CarSchema = z.object({
+  car_name: z.string(),
+  car_price: z.string(),
+  car_url: z.string(),
+  car_image: z.string(),
+  car_description: z.string(),
+});
+
+// Schema for a list of car recommendations
+const Cars = z.object({
+  cars: z.array(CarSchema),
+});
+
+// Create a function tool based on the schema
+const carRecommendationTool = zodResponsesFunction({ 
+    name: "carRecommendations", 
+    parameters: Cars 
+});
+
+// Use the tool in your OpenAI request
+const response = await openAIClient.responses.create({
+    model: "gpt-4o",
+    tools: [{ type: "web_search_preview" }, carRecommendationTool],
+    input: `${getMemoryString(relevantMemories)}\n${userInput}`,
+});
+```
+
+### Using Web Search
+
+Combine memory with web search for up-to-date recommendations:
+
+```javascript
+const response = await openAIClient.responses.create({
+    model: "gpt-4o",
+    tools: [{ type: "web_search_preview" }, carRecommendationTool],
+    input: `${getMemoryString(relevantMemories)}\n${userInput}`,
+});
+```
+
+## Examples
+
+### Complete Car Recommendation System
+
+```javascript
+import MemoryClient from "mem0ai";
+import { OpenAI } from "openai";
+import { zodResponsesFunction } from "openai/helpers/zod";
+import { z } from "zod";
+import dotenv from 'dotenv';
+
+dotenv.config();
+
+const mem0Config = {
+    apiKey: process.env.MEM0_API_KEY,
+    user_id: "sample-user",
+};
+
+async function run() {
+    // Responses without memories
+    console.log("\n\nRESPONSES WITHOUT MEMORIES\n\n");
+    await main();
+
+    // Adding sample memories
+    await addSampleMemories();
+
+    // Responses with memories
+    console.log("\n\nRESPONSES WITH MEMORIES\n\n");
+    await main(true);
+}
+
+// OpenAI Response Schema
+const CarSchema = z.object({
+  car_name: z.string(),
+  car_price: z.string(),
+  car_url: z.string(),
+  car_image: z.string(),
+  car_description: z.string(),
+});
+
+const Cars = z.object({
+  cars: z.array(CarSchema),
+});
+
+async function main(memory = false) {
+  const openAIClient = new OpenAI();
+  const mem0Client = new MemoryClient(mem0Config);
+
+  const input = "Suggest me some cars that I can buy today.";
+
+  const tool = zodResponsesFunction({ name: "carRecommendations", parameters: Cars });
+
+  // Store the user input as a memory
+  await mem0Client.add([{
+    role: "user",
+    content: input,
+  }], mem0Config);
+
+  // Search for relevant memories
+  let relevantMemories = []
+  if (memory) {
+    relevantMemories = await mem0Client.search(input, mem0Config);
+  }
+
+  const response = await openAIClient.responses.create({
+    model: "gpt-4o",
+    tools: [{ type: "web_search_preview" }, tool],
+    input: `${getMemoryString(relevantMemories)}\n${input}`,
+  });
+
+  console.log(response.output);
+}
+
+async function addSampleMemories() {
+  const mem0Client = new MemoryClient(mem0Config);
+
+  const myInterests = "I Love BMW, Audi and Porsche. I Hate Mercedes. I love Red cars and Maroon cars. I have a budget of 120K to 150K USD. I like Audi the most.";
+  
+  await mem0Client.add([{
+    role: "user",
+    content: myInterests,
+  }], mem0Config);
+}
+
+const getMemoryString = (memories) => {
+    const MEMORY_STRING_PREFIX = "These are the memories I have stored. Give more weightage to the question by users and try to answer that first. You have to modify your answer based on the memories I have provided. If the memories are irrelevant you can ignore them. Also don't reply to this section of the prompt, or the memories, they are only for your reference. The MEMORIES of the USER are: \n\n";
+    const memoryString = memories.map((mem) => `${mem.memory}`).join("\n") ?? "";
+    return memoryString.length > 0 ? `${MEMORY_STRING_PREFIX}${memoryString}` : "";
+};
+
+run().catch(console.error);
+```
+
+### Responses
+
+<CodeGroup>
+    ```json Without Memories
+    {
+      "cars": [
+        {
+          "car_name": "Toyota Camry",
+          "car_price": "$25,000",
+          "car_url": "https://www.toyota.com/camry/",
+          "car_image": "https://link-to-toyota-camry-image.com",
+          "car_description": "Reliable mid-size sedan with great fuel efficiency."
+        },
+        {
+          "car_name": "Honda Accord",
+          "car_price": "$26,000",
+          "car_url": "https://www.honda.com/accord/",
+          "car_image": "https://link-to-honda-accord-image.com",
+          "car_description": "Comfortable and spacious with advanced safety features."
+        },
+        {
+          "car_name": "Ford Mustang",
+          "car_price": "$28,000",
+          "car_url": "https://www.ford.com/mustang/",
+          "car_image": "https://link-to-ford-mustang-image.com",
+          "car_description": "Iconic sports car with powerful engine options."
+        },
+        {
+          "car_name": "Tesla Model 3",
+          "car_price": "$38,000",
+          "car_url": "https://www.tesla.com/model3",
+          "car_image": "https://link-to-tesla-model3-image.com",
+          "car_description": "Electric vehicle with advanced technology and long range."
+        },
+        {
+          "car_name": "Chevrolet Equinox",
+          "car_price": "$24,000",
+          "car_url": "https://www.chevrolet.com/equinox/",
+          "car_image": "https://link-to-chevron-equinox-image.com",
+          "car_description": "Compact SUV with a spacious interior and user-friendly technology."
+        }
+      ]
+    }
+    ```
+  
+    ```json With Memories
+    {
+      "cars": [
+        {
+          "car_name": "Audi RS7",
+          "car_price": "$118,500",
+          "car_url": "https://www.audiusa.com/us/web/en/models/rs7/2023/overview.html",
+          "car_image": "https://www.audiusa.com/content/dam/nemo/us/models/rs7/my23/gallery/1920x1080_AOZ_A717_191004.jpg",
+          "car_description": "The Audi RS7 is a high-performance hatchback with a sleek design, powerful 591-hp twin-turbo V8, and luxurious interior. It's available in various colors including red."
+        },
+        {
+          "car_name": "Porsche Panamera GTS",
+          "car_price": "$129,300",
+          "car_url": "https://www.porsche.com/usa/models/panamera/panamera-models/panamera-gts/",
+          "car_image": "https://files.porsche.com/filestore/image/multimedia/noneporsche-panamera-gts-sample-m02-high/normal/8a6327c3-6c7f-4c6f-a9a8-fb9f58b21795;sP;twebp/porsche-normal.webp",
+          "car_description": "The Porsche Panamera GTS is a luxury sports sedan with a 473-hp V8 engine, exquisite handling, and available in stunning red. Balances sportiness and comfort."
+        },
+        {
+          "car_name": "BMW M5",
+          "car_price": "$105,500",
+          "car_url": "https://www.bmwusa.com/vehicles/m-models/m5/sedan/overview.html",
+          "car_image": "https://www.bmwusa.com/content/dam/bmwusa/M/m5/2023/bmw-my23-m5-sapphire-black-twilight-purple-exterior-02.jpg",
+          "car_description": "The BMW M5 is a powerhouse sedan with a 600-hp V8 engine, known for its great handling and luxury. It comes in several distinctive colors including maroon."
+        }
+      ]
+    }
+    ```
+</CodeGroup>
+
+## Resources
+
+- [Mem0 Documentation](https://docs.mem0.ai)
+- [Mem0 Dashboard](https://app.mem0.ai/dashboard)
+- [API Reference](https://docs.mem0.ai/api-reference)
+- [OpenAI Documentation](https://platform.openai.com/docs)

docs/examples/overview.mdx

@@ -58,4 +58,12 @@ Explore how **Mem0** can power real-world applications and bring personalized, i
   <Card title="Personalized Research Agent" icon="robot" href="/examples/personalized-deep-research">  
     Build a **Deep Research AI** that remembers your research goals and compiles insights from vast information sources.  
   </Card>  
+
+  <Card title="Mem0 as an Agentic Tool" icon="robot" href="/examples/mem0-agentic-tool">  
+    Integrate Mem0's memory capabilities with OpenAI's Agents SDK to create AI agents with persistent memory.  
+  </Card>  
+
+  <Card title="OpenAI Inbuilt Tools" icon="robot" href="/examples/openai-inbuilt-tools">  
+    Use Mem0's memory capabilities with OpenAI's Inbuilt Tools to create AI agents with persistent memory.  
+  </Card>  
 </CardGroup>

docs/features/contextual-add.mdx

@@ -0,0 +1,205 @@
+---
+title: Contextual Add (ADD v2)
+icon: "square-plus"
+iconType: "solid"
+---
+
+Mem0 now supports an contextual add version (v2). To use it, set `version="v2"` during the add call. The default version is v1, which is deprecated now. We recommend migrating to `v2` for new applications.
+
+## Key Differences Between v1 and v2
+
+### Version 1 (Legacy)
+In v1 (default), users needed to pass either the entire conversation history or past k messages with each new message to generate properly contextualized memories. This approach required:
+
+- Manually tracking and sending previous messages using a sliding window approach
+- Increased payload sizes as conversations grew longer, requiring careful window size management
+
+<CodeGroup>
+
+```python Python
+# First interaction
+messages1 = [
+    {"role": "user", "content": "Hi, I'm Alex and I live in San Francisco."},
+    {"role": "assistant", "content": "Hello Alex! Nice to meet you. San Francisco is a beautiful city."}
+]
+client.add(messages1, user_id="alex")
+
+# Second interaction - must include previous messages for context
+messages2 = [
+    {"role": "user", "content": "Hi, I'm Alex and I live in San Francisco."},
+    {"role": "assistant", "content": "Hello Alex! Nice to meet you. San Francisco is a beautiful city."},
+    {"role": "user", "content": "I like to eat sushi, and yesterday I went to Sunnyvale to eat sushi with my friends."},
+    {"role": "assistant", "content": "Sushi is really a tasty choice. What did you do this weekend?"}
+]
+client.add(messages2, user_id="alex")
+```
+
+```javascript JavaScript
+// First interaction
+const messages1 = [
+    {"role": "user", "content": "Hi, I'm Alex and I live in San Francisco."},
+    {"role": "assistant", "content": "Hello Alex! Nice to meet you. San Francisco is a beautiful city."}
+];
+client.add(messages1, { user_id: "alex" })
+    .then(response => console.log(response))
+    .catch(error => console.error(error));
+
+// Second interaction - must include previous messages for context
+const messages2 = [
+    {"role": "user", "content": "Hi, I'm Alex and I live in San Francisco."},
+    {"role": "assistant", "content": "Hello Alex! Nice to meet you. San Francisco is a beautiful city."},
+    {"role": "user", "content": "I like to eat sushi, and yesterday I went to Sunnyvale to eat sushi with my friends."},
+    {"role": "assistant", "content": "Sushi is really a tasty choice. What did you do this weekend?"}
+];
+client.add(messages2, { user_id: "alex" })
+    .then(response => console.log(response))
+    .catch(error => console.error(error));
+```
+
+</CodeGroup>
+
+### Version 2 (Recommended)
+In v2, Mem0 automatically manages conversation context. Users only need to send new messages, and the system will:
+
+- Automatically retrieve relevant conversation history
+- Generate properly contextualized memories
+- Reduce payload sizes and simplify integration
+
+<CodeGroup>
+
+```python Python
+# First interaction
+messages1 = [
+    {"role": "user", "content": "Hi, I'm Alex and I live in San Francisco."},
+    {"role": "assistant", "content": "Hello Alex! Nice to meet you. San Francisco is a beautiful city."}
+]
+client.add(messages1, user_id="alex", version="v2")
+
+# Second interaction - only need to send new messages
+messages2 = [
+    {"role": "user", "content": "I like to eat sushi, and yesterday I went to Sunnyvale to eat sushi with my friends."},
+    {"role": "assistant", "content": "Sushi is really a tasty choice. What did you do this weekend?"}
+]
+client.add(messages2, user_id="alex", version="v2")
+```
+
+```javascript JavaScript
+// First interaction
+const messages1 = [
+    {"role": "user", "content": "Hi, I'm Alex and I live in San Francisco."},
+    {"role": "assistant", "content": "Hello Alex! Nice to meet you. San Francisco is a beautiful city."}
+];
+client.add(messages1, { user_id: "alex", version: "v2" })
+    .then(response => console.log(response))
+    .catch(error => console.error(error));
+
+// Second interaction - only need to send new messages
+const messages2 = [
+    {"role": "user", "content": "I like to eat sushi, and yesterday I went to Sunnyvale to eat sushi with my friends."},
+    {"role": "assistant", "content": "Sushi is really a tasty choice. What did you do this weekend?"}
+];
+client.add(messages2, { user_id: "alex", version: "v2" })
+    .then(response => console.log(response))
+    .catch(error => console.error(error));
+```
+
+</CodeGroup>
+
+## Benefits of Using v2
+
+1. **Simplified Integration**: No need to track and manage conversation history
+2. **Reduced Payload Size**: Only send new messages, not the entire conversation
+3. **Improved Memory Quality**: Automatic context retrieval ensures better memory generation
+
+## Understanding ID Parameters in v2
+
+When using contextual add v2, you have different options for how to organize and retrieve memories:
+
+### Using Only `user_id`
+
+When you provide only a `user_id`:
+
+- Memories are associated with this user's long-term memory store
+- The system will automatically retrieve relevant context from all of the user's previous conversations
+- These memories persist indefinitely across all of the user's sessions
+- Ideal for maintaining persistent user information (preferences, personal details, etc.)
+
+<CodeGroup>
+
+```python Python
+# Adding to long-term user memory
+messages = [
+    {"role": "user", "content": "I'm allergic to peanuts and shellfish."},
+    {"role": "assistant", "content": "I've noted your allergies to peanuts and shellfish."}
+]
+client.add(messages, user_id="alex", version="v2")
+```
+
+```javascript JavaScript
+// Adding to long-term user memory
+const messages = [
+    {"role": "user", "content": "I'm allergic to peanuts and shellfish."},
+    {"role": "assistant", "content": "I've noted your allergies to peanuts and shellfish."}
+];
+client.add(messages, { user_id: "alex", version: "v2" })
+    .then(response => console.log(response))
+    .catch(error => console.error(error));
+```
+
+</CodeGroup>
+
+### Using `user_id` with `run_id`
+
+When you provide both `user_id` and `run_id`:
+
+- Memories are associated with a specific conversation session or interaction
+- The system will retrieve context primarily from this specific session
+- These memories are still tied to the user but are organized by the specific session
+- Ideal for maintaining context within a specific conversation flow or task
+- Helps prevent context from different conversations from interfering with each other
+
+<CodeGroup>
+
+```python Python
+# Adding to a specific conversation session
+messages = [
+    {"role": "user", "content": "For this trip to Paris, I want to focus on art museums."},
+    {"role": "assistant", "content": "Great! I'll help you plan your Paris trip with a focus on art museums."}
+]
+client.add(messages, user_id="alex", run_id="paris-trip-2024", version="v2")
+
+# Later in the same conversation session
+messages2 = [
+    {"role": "user", "content": "I'd like to visit the Louvre on Monday."},
+    {"role": "assistant", "content": "The Louvre is a great choice for Monday. Would you like information about opening hours?"}
+]
+client.add(messages2, user_id="alex", run_id="paris-trip-2024", version="v2")
+```
+
+```javascript JavaScript
+// Adding to a specific conversation session
+const messages = [
+    {"role": "user", "content": "For this trip to Paris, I want to focus on art museums."},
+    {"role": "assistant", "content": "Great! I'll help you plan your Paris trip with a focus on art museums."}
+];
+client.add(messages, { user_id: "alex", run_id: "paris-trip-2024", version: "v2" })
+    .then(response => console.log(response))
+    .catch(error => console.error(error));
+
+// Later in the same conversation session
+const messages2 = [
+    {"role": "user", "content": "I'd like to visit the Louvre on Monday."},
+    {"role": "assistant", "content": "The Louvre is a great choice for Monday. Would you like information about opening hours?"}
+];
+client.add(messages2, { user_id: "alex", run_id: "paris-trip-2024", version: "v2" })
+    .then(response => console.log(response))
+    .catch(error => console.error(error));
+```
+
+</CodeGroup>
+
+Using `run_id` helps you organize memories into logical sessions or tasks, making it easier to maintain context for specific interactions while still associating everything with the user's overall profile.
+
+If you have any questions, please feel free to reach out to us using one of the following methods:
+
+<Snippet file="get-help.mdx" />

docs/features/custom-fact-extraction-prompt.mdx

@@ -1,25 +1,25 @@
 ---
-title: Custom Prompts
-description: 'Enhance your product experience by adding custom prompts tailored to your needs'
+title: Custom Fact Extraction Prompt
+description: 'Enhance your product experience by adding custom fact extraction prompt tailored to your needs'
 icon: "pencil"
 iconType: "solid"
 ---
 
-## Introduction to Custom Prompts
+## Introduction to Custom Fact Extraction Prompt
 
-Custom prompts allow you to tailor the behavior of your Mem0 instance to specific use cases or domains. 
-By defining a custom prompt, you can control how information is extracted, processed, and stored in your memory system.
+Custom fact extraction prompt allow you to tailor the behavior of your Mem0 instance to specific use cases or domains. 
+By defining it, you can control how information is extracted from the user's message.
 
-To create an effective custom prompt:
+To create an effective custom fact extraction prompt:
 1. Be specific about the information to extract.
 2. Provide few-shot examples to guide the LLM.
 3. Ensure examples follow the format shown below.
 
-Example of a custom prompt:
+Example of a custom fact extraction prompt:
 
 <CodeGroup>
 ```python Python
-custom_prompt = """
+custom_fact_extraction_prompt = """
 Please only extract entities containing customer support information, order details, and user information. 
 Here are some few shot examples:
 
@@ -67,7 +67,7 @@ Return the facts and customer information in a json format as shown above.
 ```
 </CodeGroup>
 
-Here we initialize the custom prompt in the config:
+Here we initialize the custom fact extraction prompt in the config:
 
 <CodeGroup>
 ```python Python
@@ -82,7 +82,7 @@ config = {
             "max_tokens": 2000,
         }
     },
-    "custom_prompt": custom_prompt,
+    "custom_fact_extraction_prompt": custom_fact_extraction_prompt,
     "version": "v1.1"
 }
 
@@ -166,4 +166,4 @@ await memory.add('I like going to hikes', { userId: "user123" });
 ```
 </CodeGroup>
 
-The custom prompt will process both the user and assistant messages to extract relevant information according to the defined format.
+The custom fact extraction prompt will process both the user and assistant messages to extract relevant information according to the defined format.

docs/features/custom-update-memory-prompt.mdx

@@ -0,0 +1,239 @@
+---
+title: Custom Update Memory Prompt
+icon: "pencil"
+iconType: "solid"
+---
+Update memory prompt is a prompt used to determine the action to be performed on the memory. 
+By customizing this prompt, you can control how the memory is updated.  
+  
+
+## Introduction  
+Mem0 memory system compares the newly retrieved facts with the existing memory and determines the action to be performed on the memory.  
+The kinds of actions are:
+- Add
+    - Add the newly retrieved facts to the memory.
+- Update
+    - Update the existing memory with the newly retrieved facts.
+- Delete
+    - Delete the existing memory.
+- No Change
+    - Do not make any changes to the memory.
+  
+### Example
+Example of a custom update memory prompt:
+
+<CodeGroup>
+```python Python
+UPDATE_MEMORY_PROMPT = """You are a smart memory manager which controls the memory of a system.
+You can perform four operations: (1) add into the memory, (2) update the memory, (3) delete from the memory, and (4) no change.
+
+Based on the above four operations, the memory will change.
+
+Compare newly retrieved facts with the existing memory. For each new fact, decide whether to:
+- ADD: Add it to the memory as a new element
+- UPDATE: Update an existing memory element
+- DELETE: Delete an existing memory element
+- NONE: Make no change (if the fact is already present or irrelevant)
+
+There are specific guidelines to select which operation to perform:
+
+1. **Add**: If the retrieved facts contain new information not present in the memory, then you have to add it by generating a new ID in the id field.
+- **Example**:
+    - Old Memory:
+        [
+            {
+                "id" : "0",
+                "text" : "User is a software engineer"
+            }
+        ]
+    - Retrieved facts: ["Name is John"]
+    - New Memory:
+        {
+            "memory" : [
+                {
+                    "id" : "0",
+                    "text" : "User is a software engineer",
+                    "event" : "NONE"
+                },
+                {
+                    "id" : "1",
+                    "text" : "Name is John",
+                    "event" : "ADD"
+                }
+            ]
+
+        }
+
+2. **Update**: If the retrieved facts contain information that is already present in the memory but the information is totally different, then you have to update it. 
+If the retrieved fact contains information that conveys the same thing as the elements present in the memory, then you have to keep the fact which has the most information. 
+Example (a) -- if the memory contains "User likes to play cricket" and the retrieved fact is "Loves to play cricket with friends", then update the memory with the retrieved facts.
+Example (b) -- if the memory contains "Likes cheese pizza" and the retrieved fact is "Loves cheese pizza", then you do not need to update it because they convey the same information.
+If the direction is to update the memory, then you have to update it.
+Please keep in mind while updating you have to keep the same ID.
+Please note to return the IDs in the output from the input IDs only and do not generate any new ID.
+- **Example**:
+    - Old Memory:
+        [
+            {
+                "id" : "0",
+                "text" : "I really like cheese pizza"
+            },
+            {
+                "id" : "1",
+                "text" : "User is a software engineer"
+            },
+            {
+                "id" : "2",
+                "text" : "User likes to play cricket"
+            }
+        ]
+    - Retrieved facts: ["Loves chicken pizza", "Loves to play cricket with friends"]
+    - New Memory:
+        {
+        "memory" : [
+                {
+                    "id" : "0",
+                    "text" : "Loves cheese and chicken pizza",
+                    "event" : "UPDATE",
+                    "old_memory" : "I really like cheese pizza"
+                },
+                {
+                    "id" : "1",
+                    "text" : "User is a software engineer",
+                    "event" : "NONE"
+                },
+                {
+                    "id" : "2",
+                    "text" : "Loves to play cricket with friends",
+                    "event" : "UPDATE",
+                    "old_memory" : "User likes to play cricket"
+                }
+            ]
+        }
+
+
+3. **Delete**: If the retrieved facts contain information that contradicts the information present in the memory, then you have to delete it. Or if the direction is to delete the memory, then you have to delete it.
+Please note to return the IDs in the output from the input IDs only and do not generate any new ID.
+- **Example**:
+    - Old Memory:
+        [
+            {
+                "id" : "0",
+                "text" : "Name is John"
+            },
+            {
+                "id" : "1",
+                "text" : "Loves cheese pizza"
+            }
+        ]
+    - Retrieved facts: ["Dislikes cheese pizza"]
+    - New Memory:
+        {
+        "memory" : [
+                {
+                    "id" : "0",
+                    "text" : "Name is John",
+                    "event" : "NONE"
+                },
+                {
+                    "id" : "1",
+                    "text" : "Loves cheese pizza",
+                    "event" : "DELETE"
+                }
+        ]
+        }
+
+4. **No Change**: If the retrieved facts contain information that is already present in the memory, then you do not need to make any changes.
+- **Example**:
+    - Old Memory:
+        [
+            {
+                "id" : "0",
+                "text" : "Name is John"
+            },
+            {
+                "id" : "1",
+                "text" : "Loves cheese pizza"
+            }
+        ]
+    - Retrieved facts: ["Name is John"]
+    - New Memory:
+        {
+        "memory" : [
+                {
+                    "id" : "0",
+                    "text" : "Name is John",
+                    "event" : "NONE"
+                },
+                {
+                    "id" : "1",
+                    "text" : "Loves cheese pizza",
+                    "event" : "NONE"
+                }
+            ]
+        }
+"""
+```
+</CodeGroup>  
+
+## Output format
+The prompt needs to guide the output to follow the structure as shown below:
+<CodeGroup>
+```json Add
+{
+    "memory": [
+        {
+            "id" : "0",
+            "text" : "This information is new",
+            "event" : "ADD"
+        }
+    ]
+}
+```
+
+```json Update
+{
+    "memory": [
+        {
+            "id" : "0",
+            "text" : "This information replaces the old information",
+            "event" : "UPDATE",
+            "old_memory" : "Old information"
+        }
+    ]
+}
+```
+
+```json Delete
+{
+    "memory": [
+        {
+            "id" : "0",
+            "text" : "This information will be deleted",
+            "event" : "DELETE"
+        }
+    ]
+}
+```
+
+```json No Change
+{
+    "memory": [
+        {
+            "id" : "0",
+            "text" : "No changes for this information",
+            "event" : "NONE"
+        }
+    ]
+}
+```
+</CodeGroup>
+  
+  
+## custom update memory prompt vs custom prompt
+
+| Feature | `custom_update_memory_prompt` | `custom_prompt` |
+|---------|-------------------------------|-----------------|
+| Use case | Determine the action to be performed on the memory | Extract the facts from messages |
+| Reference | Retrieved facts from messages and old memory | Messages |
+| Output | Action to be performed on the memory | Extracted facts |

docs/features/memory-export.mdx

@@ -71,13 +71,32 @@ Here's an example schema for extracting professional profile information:
 
 ### Submit Export Job
 
+You can optionally provide additional instructions to guide how memories are processed and structured during export using the `export_instructions` parameter.
+
 <CodeGroup>
 
 ```python Python
+# Basic export request
 response = client.create_memory_export(
     schema=json_schema,
-    user_id="user123"
+    user_id="alice"
 )
+
+# Export with custom instructions
+export_instructions = """
+1. Create a comprehensive profile with detailed information in each category
+2. Only mark fields as "None" when absolutely no relevant information exists
+3. Base all information directly on the user's memories
+4. When contradictions exist, prioritize the most recent information
+5. Clearly distinguish between factual statements and inferences
+"""
+
+response = client.create_memory_export(
+    schema=json_schema,
+    user_id="alice",
+    export_instructions=export_instructions
+)
+
 print(response)
 ```
 
@@ -87,7 +106,8 @@ curl -X POST "https://api.mem0.ai/v1/memories/export/" \
      -H "Content-Type: application/json" \
      -d '{
          "schema": {json_schema},
-         "user_id": "user123"
+         "user_id": "alice",
+         "export_instructions": "1. Create a comprehensive profile with detailed information\n2. Only mark fields as \"None\" when absolutely no relevant information exists"
      }'
 ```
 
@@ -107,12 +127,12 @@ Once the export job is complete, you can retrieve the structured data:
 <CodeGroup>
 
 ```python Python
-response = client.get_memory_export(user_id="user123")
+response = client.get_memory_export(user_id="alice")
 print(response)
 ```
 
 ```bash cURL
-curl -X GET "https://api.mem0.ai/v1/memories/export/?user_id=user123" \
+curl -X GET "https://api.mem0.ai/v1/memories/export/?user_id=alice" \
      -H "Authorization: Token your-api-key"
 ```
 

docs/features/platform-overview.mdx

@@ -12,6 +12,9 @@ Learn about the key features and capabilities that make Mem0 a powerful platform
   <Card title="Advanced Retrieval" icon="magnifying-glass" href="/features/advanced-retrieval">
     Superior search results using state-of-the-art algorithms, including keyword search, reranking, and filtering capabilities.
   </Card>
+  <Card title="Contextual Add" icon="square-plus" href="/features/contextual-add">
+    Only send your latest conversation history - we automatically retrieve the rest and generate properly contextualized memories.
+  </Card>
   <Card title="Multimodal Support" icon="photo-film" href="/features/multimodal-support">
     Process and analyze various types of content including images.
   </Card>

docs/integrations/langchain-tools.mdx

@@ -95,7 +95,7 @@ add_input = {
         {"role": "user", "content": "Hi, I'm Alex. I'm a vegetarian and I'm allergic to nuts."},
         {"role": "assistant", "content": "Hello Alex! I've noted that you're a vegetarian and have a nut allergy."}
     ],
-    "user_id": "alex123",
+    "user_id": "alex",
     "output_format": "v1.1",
     "metadata": {"food": "vegan"}
 }
@@ -173,7 +173,7 @@ search_input = {
     "filters": {
         "AND": [
             {"created_at": {"gte": "2024-07-20", "lte": "2024-12-10"}},
-            {"user_id": "alex123"}
+            {"user_id": "alex"}
         ]
     },
     "version": "v2"
@@ -186,7 +186,7 @@ result = search_tool.invoke(search_input)
   {
     "id": "1a75e827-7eca-45ea-8c5c-cfd43299f061",
     "memory": "Name is Alex",
-    "user_id": "alex123", 
+    "user_id": "alex", 
     "hash": "d0fccc8fa47f7a149ee95750c37bb0ca",
     "metadata": {
       "food": "vegan"
@@ -255,7 +255,7 @@ get_all_input = {
     "version": "v2",
     "filters": {
         "AND": [
-            {"user_id": "alex123"},
+            {"user_id": "alex"},
             {"created_at": {"gte": "2024-07-01", "lte": "2024-12-31"}}
         ]
     },
@@ -274,7 +274,7 @@ get_all_result = get_all_tool.invoke(get_all_input)
     {
       "id": "1a75e827-7eca-45ea-8c5c-cfd43299f061",
       "memory": "Name is Alex",
-      "user_id": "alex123", 
+      "user_id": "alex", 
       "hash": "d0fccc8fa47f7a149ee95750c37bb0ca",
       "metadata": {
         "food": "vegan"
@@ -288,7 +288,7 @@ get_all_result = get_all_tool.invoke(get_all_input)
     {
       "id": "91509588-0b39-408a-8df3-84b3bce8c521",
       "memory": "Is a vegetarian",
-      "user_id": "alex123",
+      "user_id": "alex",
       "hash": "ce6b1c84586772ab9995a9477032df99", 
       "metadata": {
         "food": "vegan"
@@ -303,7 +303,7 @@ get_all_result = get_all_tool.invoke(get_all_input)
     {
       "id": "8d74f7a0-6107-4589-bd6f-210f6bf4fbbb",
       "memory": "Is allergic to nuts",
-      "user_id": "alex123",
+      "user_id": "alex",
       "hash": "7873cd0e5a29c513253d9fad038e758b",
       "metadata": {
         "food": "vegan"

docs/open-source/python-quickstart.mdx

@@ -85,14 +85,18 @@ m = Memory.from_config(config_dict=config)
 
 <CodeGroup>
 ```python Code
-const messages = [
+messages = [
     {"role": "user", "content": "I'm planning to watch a movie tonight. Any recommendations?"},
     {"role": "assistant", "content": "How about a thriller movies? They can be quite engaging."},
     {"role": "user", "content": "I'm not a big fan of thriller movies but I love sci-fi movies."},
     {"role": "assistant", "content": "Got it! I'll avoid thriller recommendations and suggest sci-fi movies in the future."}
 ]
 
+# Store inferred memories (default behavior)
 result = m.add(messages, user_id="alice", metadata={"category": "movie_recommendations"})
+
+# Store raw messages without inference
+# result = m.add(messages, user_id="alice", metadata={"category": "movie_recommendations"}, infer=False)
 ```
 
 ```json Output
@@ -368,7 +372,8 @@ Mem0 offers extensive configuration options to customize its behavior according
 |------------------|--------------------------------------|----------------------------|
 | `history_db_path` | Path to the history database         | "{mem0_dir}/history.db"    |
 | `version`         | API version                          | "v1.1"                     |
-| `custom_prompt`   | Custom prompt for memory processing  | None                       |
+| `custom_fact_extraction_prompt`   | Custom prompt for memory processing  | None                       |
+| `custom_update_memory_prompt` | Custom prompt for update memory | None                |
 </Accordion>
 
 <Accordion title="Complete Configuration Example">
@@ -405,7 +410,8 @@ config = {
     },
     "history_db_path": "/path/to/history.db",
     "version": "v1.1",
-    "custom_prompt": "Optional custom prompt for memory processing"
+    "custom_fact_extraction_prompt": "Optional custom prompt for fact extraction for memory",
+    "custom_update_memory_prompt": "Optional custom prompt for update memory"
 }
 ```
 </Accordion>

docs/openapi.json

@@ -932,27 +932,27 @@
 				"x-code-samples": [
 				{
 				"lang": "Python",
-				"source": "# To use the Python SDK, install the package:\n# pip install mem0ai\n\nfrom mem0 import MemoryClient\n\nclient = MemoryClient(api_key=\"your_api_key\", org_id=\"your_org_id\", project_id=\"your_project_id\")\n\nmessages = [\n    {\"role\": \"user\", \"content\": \"<user-message>\"},\n    {\"role\": \"assistant\", \"content\": \"<assistant-response>\"}\n]\n\nclient.add(messages, user_id=\"<user-id>\")"
+				"source": "# To use the Python SDK, install the package:\n# pip install mem0ai\n\nfrom mem0 import MemoryClient\n\nclient = MemoryClient(api_key=\"your_api_key\", org_id=\"your_org_id\", project_id=\"your_project_id\")\n\nmessages = [\n    {\"role\": \"user\", \"content\": \"<user-message>\"},\n    {\"role\": \"assistant\", \"content\": \"<assistant-response>\"}\n]\n\nclient.add(messages, user_id=\"<user-id>\", version=\"v2\")"
 				},
 				{
 				"lang": "JavaScript",
-				"source": "// To use the JavaScript SDK, install the package:\n// npm i mem0ai\n\nimport MemoryClient from 'mem0ai';\nconst client = new MemoryClient({ apiKey: \"your-api-key\" });\n\nconst messages = [\n  { role: \"user\", content: \"Hi, I'm Alex. I'm a vegetarian and I'm allergic to nuts.\" },\n  { role: \"assistant\", content: \"Hello Alex! I've noted that you're a vegetarian and have a nut allergy. I'll keep this in mind for any food-related recommendations or discussions.\" }\n];\n\nclient.add(messages, { user_id: \"<user_id>\" })\n  .then(result => console.log(result))\n  .catch(error => console.error(error));"
+				"source": "// To use the JavaScript SDK, install the package:\n// npm i mem0ai\n\nimport MemoryClient from 'mem0ai';\nconst client = new MemoryClient({ apiKey: \"your-api-key\" });\n\nconst messages = [\n  { role: \"user\", content: \"Hi, I'm Alex. I'm a vegetarian and I'm allergic to nuts.\" },\n  { role: \"assistant\", content: \"Hello Alex! I've noted that you're a vegetarian and have a nut allergy. I'll keep this in mind for any food-related recommendations or discussions.\" }\n];\n\nclient.add(messages, { user_id: \"<user_id>\", version: \"v2\" })\n  .then(result => console.log(result))\n  .catch(error => console.error(error));"
 				},				  
 				{
 				"lang": "cURL",
-				"source": "curl --request POST \\\n  --url https://api.mem0.ai/v1/memories/ \\\n  --header 'Authorization: Token <api-key>' \\\n  --header 'Content-Type: application/json' \\\n  --data '{\n  \"messages\": [\n    {}\n  ],\n  \"agent_id\": \"<string>\",\n  \"user_id\": \"<string>\",\n  \"app_id\": \"<string>\",\n  \"run_id\": \"<string>\",\n  \"metadata\": {},\n  \"includes\": \"<string>\",\n  \"excludes\": \"<string>\",\n  \"infer\": true,\n  \"custom_categories\": {}, \n  \"org_id\": \"<string>\",\n  \"project_id\": \"<string>\"\n}'"
+				"source": "curl --request POST \\\n  --url https://api.mem0.ai/v1/memories/ \\\n  --header 'Authorization: Token <api-key>' \\\n  --header 'Content-Type: application/json' \\\n  --data '{\n  \"messages\": [\n    {}\n  ],\n  \"agent_id\": \"<string>\",\n  \"user_id\": \"<string>\",\n  \"app_id\": \"<string>\",\n  \"run_id\": \"<string>\",\n  \"metadata\": {},\n  \"includes\": \"<string>\",\n  \"excludes\": \"<string>\",\n  \"infer\": true,\n  \"custom_categories\": {}, \n  \"org_id\": \"<string>\",\n  \"project_id\": \"<string>\",\n  \"version\": \"v2\"\n}'"
 				},
 				{
 				"lang": "Go",
-				"source": "package main\n\nimport (\n\t\"fmt\"\n\t\"strings\"\n\t\"net/http\"\n\t\"io/ioutil\"\n)\n\nfunc main() {\n\n\turl := \"https://api.mem0.ai/v1/memories/\"\n\n\tpayload := strings.NewReader(\"{\n  \\\"messages\\\": [\n    {}\n  ],\n  \\\"agent_id\\\": \\\"<string>\\\",\n  \\\"user_id\\\": \\\"<string>\\\",\n  \\\"app_id\\\": \\\"<string>\\\",\n  \\\"run_id\\\": \\\"<string>\\\",\n  \\\"metadata\\\": {},\n  \\\"includes\\\": \\\"<string>\\\",\n  \\\"excludes\\\": \\\"<string>\\\",\n  \\\"infer\\\": true,\n  \\\"custom_categories\\\": {},\n  \\\"org_id\\\": \\\"<string>\\\",\n  \\\"project_id\\\": \\\"<string>\"\n}\")\n\n\treq, _ := http.NewRequest(\"POST\", url, payload)\n\n\treq.Header.Add(\"Authorization\", \"Token <api-key>\")\n\treq.Header.Add(\"Content-Type\", \"application/json\")\n\n\tres, _ := http.DefaultClient.Do(req)\n\n\tdefer res.Body.Close()\n\tbody, _ := ioutil.ReadAll(res.Body)\n\n\tfmt.Println(res)\n\tfmt.Println(string(body))\n\n}"
+				"source": "package main\n\nimport (\n\t\"fmt\"\n\t\"strings\"\n\t\"net/http\"\n\t\"io/ioutil\"\n)\n\nfunc main() {\n\n\turl := \"https://api.mem0.ai/v1/memories/\"\n\n\tpayload := strings.NewReader(\"{\n  \\\"messages\\\": [\n    {}\n  ],\n  \\\"agent_id\\\": \\\"<string>\\\",\n  \\\"user_id\\\": \\\"<string>\\\",\n  \\\"app_id\\\": \\\"<string>\\\",\n  \\\"run_id\\\": \\\"<string>\\\",\n  \\\"metadata\\\": {},\n  \\\"includes\\\": \\\"<string>\\\",\n  \\\"excludes\\\": \\\"<string>\\\",\n  \\\"infer\\\": true,\n  \\\"custom_categories\\\": {},\n  \\\"org_id\\\": \\\"<string>\\\",\n  \\\"project_id\\\": \\\"<string>\",\n  \\\"version\\\": \"v2\"\n}\")\n\n\treq, _ := http.NewRequest(\"POST\", url, payload)\n\n\treq.Header.Add(\"Authorization\", \"Token <api-key>\")\n\treq.Header.Add(\"Content-Type\", \"application/json\")\n\n\tres, _ := http.DefaultClient.Do(req)\n\n\tdefer res.Body.Close()\n\tbody, _ := ioutil.ReadAll(res.Body)\n\n\tfmt.Println(res)\n\tfmt.Println(string(body))\n\n}"
 				},
 				{
 				"lang": "PHP",
-				"source": "<?php\n\n$curl = curl_init();\n\ncurl_setopt_array($curl, [\n  CURLOPT_URL => \"https://api.mem0.ai/v1/memories/\",\n  CURLOPT_RETURNTRANSFER => true,\n  CURLOPT_ENCODING => \"\",\n  CURLOPT_MAXREDIRS => 10,\n  CURLOPT_TIMEOUT => 30,\n  CURLOPT_HTTP_VERSION => CURL_HTTP_VERSION_1_1,\n  CURLOPT_CUSTOMREQUEST => \"POST\",\n  CURLOPT_POSTFIELDS => \"{\n  \\\"messages\\\": [\n    {}\n  ],\n  \\\"agent_id\\\": \\\"<string>\\\",\n  \\\"user_id\\\": \\\"<string>\\\",\n  \\\"app_id\\\": \\\"<string>\\\",\n  \\\"run_id\\\": \\\"<string>\\\",\n  \\\"metadata\\\": {},\n  \\\"includes\\\": \\\"<string>\\\",\n  \\\"excludes\\\": \\\"<string>\\\",\n  \\\"infer\\\": true,\n  \\\"custom_categories\\\": {}, \n  \\\"org_id\\\": \\\"<string>\\\",\n  \\\"project_id\\\": \\\"<string>\"\n}\",\n  CURLOPT_HTTPHEADER => [\n    \"Authorization: Token <api-key>\",\n    \"Content-Type: application/json\"\n  ],\n]);\n\n$response = curl_exec($curl);\n$err = curl_error($curl);\n\ncurl_close($curl);\n\nif ($err) {\n  echo \"cURL Error #:\" . $err;\n} else {\n  echo $response;\n}"
+				"source": "<?php\n\n$curl = curl_init();\n\ncurl_setopt_array($curl, [\n  CURLOPT_URL => \"https://api.mem0.ai/v1/memories/\",\n  CURLOPT_RETURNTRANSFER => true,\n  CURLOPT_ENCODING => \"\",\n  CURLOPT_MAXREDIRS => 10,\n  CURLOPT_TIMEOUT => 30,\n  CURLOPT_HTTP_VERSION => CURL_HTTP_VERSION_1_1,\n  CURLOPT_CUSTOMREQUEST => \"POST\",\n  CURLOPT_POSTFIELDS => \"{\n  \\\"messages\\\": [\n    {}\n  ],\n  \\\"agent_id\\\": \\\"<string>\\\",\n  \\\"user_id\\\": \\\"<string>\\\",\n  \\\"app_id\\\": \\\"<string>\\\",\n  \\\"run_id\\\": \\\"<string>\\\",\n  \\\"metadata\\\": {},\n  \\\"includes\\\": \\\"<string>\\\",\n  \\\"excludes\\\": \\\"<string>\\\",\n  \\\"infer\\\": true,\n  \\\"custom_categories\\\": {}, \n  \\\"org_id\\\": \\\"<string>\\\",\n  \\\"project_id\\\": \\\"<string>\",\n  \\\"version\\\": \"v2\"\n}\",\n  CURLOPT_HTTPHEADER => [\n    \"Authorization: Token <api-key>\",\n    \"Content-Type: application/json\"\n  ],\n]);\n\n$response = curl_exec($curl);\n$err = curl_error($curl);\n\ncurl_close($curl);\n\nif ($err) {\n  echo \"cURL Error #:\" . $err;\n} else {\n  echo $response;\n}"
 				},
 				{
 				"lang": "Java",
-				"source": "HttpResponse<String> response = Unirest.post(\"https://api.mem0.ai/v1/memories/\")\n  .header(\"Authorization\", \"Token <api-key>\")\n  .header(\"Content-Type\", \"application/json\")\n  .body(\"{\n  \\\"messages\\\": [\n    {}\n  ],\n  \\\"agent_id\\\": \\\"<string>\\\",\n  \\\"user_id\\\": \\\"<string>\\\",\n  \\\"app_id\\\": \\\"<string>\\\",\n  \\\"run_id\\\": \\\"<string>\\\",\n  \\\"metadata\\\": {},\n  \\\"includes\\\": \\\"<string>\\\",\n  \\\"excludes\\\": \\\"<string>\\\",\n  \\\"infer\\\": true,\n  \\\"custom_categories\\\": {}, \n  \\\"org_id\\\": \\\"<string>\\\",\n  \\\"project_id\\\": \\\"<string>\"\n}\")\n  .asString();"
+				"source": "HttpResponse<String> response = Unirest.post(\"https://api.mem0.ai/v1/memories/\")\n  .header(\"Authorization\", \"Token <api-key>\")\n  .header(\"Content-Type\", \"application/json\")\n  .body(\"{\n  \\\"messages\\\": [\n    {}\n  ],\n  \\\"agent_id\\\": \\\"<string>\\\",\n  \\\"user_id\\\": \\\"<string>\\\",\n  \\\"app_id\\\": \\\"<string>\\\",\n  \\\"run_id\\\": \\\"<string>\\\",\n  \\\"metadata\\\": {},\n  \\\"includes\\\": \\\"<string>\\\",\n  \\\"excludes\\\": \\\"<string>\\\",\n  \\\"infer\\\": true,\n  \\\"custom_categories\\\": {}, \n  \\\"org_id\\\": \\\"<string>\\\",\n  \\\"project_id\\\": \\\"<string>\",\n  \\\"version\\\": \"v2\"\n}\")\n  .asString();"
 					}
 				],
 				"x-codegen-request-body-name": "data"
@@ -4809,6 +4809,12 @@
 						"title": "Project id",
 						"type": "string",
 						"nullable": true
+					},
+					"version": {
+						"description": "The version of the memory to use. The default version is v1, which is deprecated. We recommend using v2 for new applications.",
+						"title": "Version", 
+						"type": "string",
+						"nullable": true
 					}
 				}
 			},

docs/platform/quickstart.mdx

@@ -87,10 +87,10 @@ messages = [
 ]
 
 # The default output_format is v1.0
-client.add(messages, user_id="alex", output_format="v1.0")
+client.add(messages, user_id="alex", output_format="v1.0", version="v2")
 
 # To use the latest output_format, set the output_format parameter to "v1.1"
-client.add(messages, user_id="alex", output_format="v1.1", metadata={"food": "vegan"})
+client.add(messages, user_id="alex", output_format="v1.1", metadata={"food": "vegan"}, version="v2")
 ```
 
 ```javascript JavaScript
@@ -98,7 +98,7 @@ const messages = [
     {"role": "user", "content": "Hi, I'm Alex. I'm a vegetarian and I'm allergic to nuts."},
     {"role": "assistant", "content": "Hello Alex! I've noted that you're a vegetarian and have a nut allergy. I'll keep this in mind for any food-related recommendations or discussions."}
 ];
-client.add(messages, { user_id: "alex", output_format: "v1.1", metadata: { food: "vegan" } })
+client.add(messages, { user_id: "alex", output_format: "v1.1", metadata: { food: "vegan" }, version: "v2" })
     .then(response => console.log(response))
     .catch(error => console.error(error));
 ```
@@ -116,7 +116,8 @@ curl -X POST "https://api.mem0.ai/v1/memories/" \
          "output_format": "v1.1",
          "metadata": {
              "food": "vegan"
-         }
+         },
+         "version": "v2"
      }'
 ```
 
@@ -191,10 +192,10 @@ messages = [
 ]
 
 # The default output_format is v1.0
-client.add(messages, user_id="alex123", run_id="trip-planning-2024", output_format="v1.0")
+client.add(messages, user_id="alex", run_id="trip-planning-2024", output_format="v1.0", version="v2")
 
 # To use the latest output_format, set the output_format parameter to "v1.1"
-client.add(messages, user_id="alex123", run_id="trip-planning-2024", output_format="v1.1")
+client.add(messages, user_id="alex", run_id="trip-planning-2024", output_format="v1.1", version="v2")
 ```
 
 ```javascript JavaScript
@@ -204,7 +205,7 @@ const messages = [
     {"role": "user", "content": "Yes, please! Especially in Tokyo."},
     {"role": "assistant", "content": "Great! I'll remember that you're interested in vegetarian restaurants in Tokyo for your upcoming trip. I'll prepare a list for you in our next interaction."}
 ];
-client.add(messages, { user_id: "alex123", run_id: "trip-planning-2024", output_format: "v1.1" })
+client.add(messages, { user_id: "alex", run_id: "trip-planning-2024", output_format: "v1.1", version: "v2" })
     .then(response => console.log(response))
     .catch(error => console.error(error));
 ```
@@ -220,9 +221,10 @@ curl -X POST "https://api.mem0.ai/v1/memories/" \
              {"role": "user", "content": "Yes, please! Especially in Tokyo."},
              {"role": "assistant", "content": "Great! I'll remember that you're interested in vegetarian restaurants in Tokyo for your upcoming trip. I'll prepare a list for you in our next interaction."}
          ],
-         "user_id": "alex123",
+         "user_id": "alex",
          "run_id": "trip-planning-2024",
-         "output_format": "v1.1"
+         "output_format": "v1.1",
+         "version": "v2"
      }'
 ```
 
@@ -274,10 +276,10 @@ messages = [
 ]
 
 # The default output_format is v1.0
-client.add(messages, agent_id="ai-tutor", output_format="v1.0")
+client.add(messages, agent_id="ai-tutor", output_format="v1.0", version="v2")
 
 # To use the latest output_format, set the output_format parameter to "v1.1"
-client.add(messages, agent_id="ai-tutor", output_format="v1.1")
+client.add(messages, agent_id="ai-tutor", output_format="v1.1", version="v2")
 ```
 
 ```javascript JavaScript
@@ -285,7 +287,7 @@ const messages = [
     {"role": "system", "content": "You are an AI tutor with a personality. Give yourself a name for the user."},
     {"role": "assistant", "content": "Understood. I'm an AI tutor with a personality. My name is Alice."}
 ];
-client.add(messages, { agent_id: "ai-tutor", output_format: "v1.1" })
+client.add(messages, { agent_id: "ai-tutor", output_format: "v1.1", version: "v2" })
     .then(response => console.log(response))
     .catch(error => console.error(error));
 ```
@@ -300,7 +302,8 @@ curl -X POST "https://api.mem0.ai/v1/memories/" \
              {"role": "assistant", "content": "Understood. I'm an AI tutor with a personality. My name is Alice."}
          ],
          "agent_id": "ai-tutor",
-         "output_format": "v1.1"
+         "output_format": "v1.1",
+         "version": "v2"
      }'
 ```
 
@@ -368,7 +371,7 @@ messages = [
     {"role": "assistant", "content": "That's great! I'm going to Dubai next month."},
 ]
 
-client.add(messages=messages, user_id="user1", agent_id="agent1")
+client.add(messages=messages, user_id="user1", agent_id="agent1", version="v2")
 ```
 
 ```javascript JavaScript
@@ -377,7 +380,7 @@ const messages = [
     {"role": "assistant", "content": "That's great! I'm going to Dubai next month."},
 ]
 
-client.add(messages, { user_id: "user1", agent_id: "agent1" })
+client.add(messages, { user_id: "user1", agent_id: "agent1", version: "v2" })
     .then(response => console.log(response))
     .catch(error => console.error(error));
 ```
@@ -392,7 +395,8 @@ curl -X POST "https://api.mem0.ai/v1/memories/" \
              {"role": "assistant", "content": "That's great! I'm going to Dubai next month."},
          ],
          "user_id": "user1",
-         "agent_id": "agent1"
+         "agent_id": "agent1",
+         "version": "v2"
      }'
 ```
 
@@ -1004,17 +1008,17 @@ curl -X GET "https://api.mem0.ai/v1/memories/?agent_id=ai-tutor&page=1&page_size
 <CodeGroup>
 
 ```python Python
-short_term_memories = client.get_all(user_id="alex123", run_id="trip-planning-2024", page=1, page_size=50)
+short_term_memories = client.get_all(user_id="alex", run_id="trip-planning-2024", page=1, page_size=50)
 ```
 
 ```javascript JavaScript
-client.getAll({ user_id: "alex123", run_id: "trip-planning-2024", page: 1, page_size: 50 })
+client.getAll({ user_id: "alex", run_id: "trip-planning-2024", page: 1, page_size: 50 })
     .then(memories => console.log(memories))
     .catch(error => console.error(error));
 ```
 
 ```bash cURL
-curl -X GET "https://api.mem0.ai/v1/memories/?user_id=alex123&run_id=trip-planning-2024&page=1&page_size=50" \
+curl -X GET "https://api.mem0.ai/v1/memories/?user_id=alex&run_id=trip-planning-2024&page=1&page_size=50" \
      -H "Authorization: Token your-api-key"
 ```
 
@@ -1027,7 +1031,7 @@ curl -X GET "https://api.mem0.ai/v1/memories/?user_id=alex123&run_id=trip-planni
         {
             "id":"06d8df63-7bd2-4fad-9acb-60871bcecee0",
             "memory":"Planning a trip to Japan next month. Interested in vegetarian restaurants in Tokyo.",
-            "user_id":"alex123",
+            "user_id":"alex",
             "hash":"d2088c936e259f2f5d2d75543d31401c",
             "metadata":None,
             "immutable": false,
@@ -1038,7 +1042,7 @@ curl -X GET "https://api.mem0.ai/v1/memories/?user_id=alex123&run_id=trip-planni
         {
             "id":"b4229775-d860-4ccb-983f-0f628ca112f5",
             "memory":"Planning a trip to Japan next month. Interested in vegetarian restaurants in Tokyo.",
-            "user_id":"alex123",
+            "user_id":"alex",
             "hash":"d2088c936e259f2f5d2d75543d31401c",
             "metadata":None,
             "immutable": false,
@@ -1049,7 +1053,7 @@ curl -X GET "https://api.mem0.ai/v1/memories/?user_id=alex123&run_id=trip-planni
         {
             "id":"df1aca24-76cf-4b92-9f58-d03857efcb64",
             "memory":"Planning a trip to Japan next month. Interested in vegetarian restaurants in Tokyo.",
-            "user_id":"alex123",
+            "user_id":"alex",
             "hash":"d2088c936e259f2f5d2d75543d31401c",
             "metadata":None,
             "immutable": false,
@@ -1072,7 +1076,7 @@ curl -X GET "https://api.mem0.ai/v1/memories/?user_id=alex123&run_id=trip-planni
             {
                 "id": "06d8df63-7bd2-4fad-9acb-60871bcecee0",
                 "memory": "Planning a trip to Japan next month. Interested in vegetarian restaurants in Tokyo.",
-                "user_id": "alex123",
+                "user_id": "alex",
                 "hash": "d2088c936e259f2f5d2d75543d31401c",
                 "metadata":None,
                 "immutable": false,
@@ -1083,7 +1087,7 @@ curl -X GET "https://api.mem0.ai/v1/memories/?user_id=alex123&run_id=trip-planni
             {
                 "id": "b4229775-d860-4ccb-983f-0f628ca112f5",
                 "memory": "Planning a trip to Japan next month. Interested in vegetarian restaurants in Tokyo.",
-                "user_id": "alex123",
+                "user_id": "alex",
                 "hash": "d2088c936e259f2f5d2d75543d31401c",
                 "metadata":None,
                 "immutable": false,
@@ -1094,7 +1098,7 @@ curl -X GET "https://api.mem0.ai/v1/memories/?user_id=alex123&run_id=trip-planni
             {
                 "id": "df1aca24-76cf-4b92-9f58-d03857efcb64",
                 "memory": "Planning a trip to Japan next month. Interested in vegetarian restaurants in Tokyo.",
-                "user_id": "alex123",
+                "user_id": "alex",
                 "hash": "d2088c936e259f2f5d2d75543d31401c",
                 "metadata":None,
                 "immutable": false,
@@ -1133,7 +1137,7 @@ curl -X GET "https://api.mem0.ai/v1/memories/582bbe6d-506b-48c6-a4c6-5df3b1e6342
 {
    "id":"06d8df63-7bd2-4fad-9acb-60871bcecee0",
    "memory":"Planning a trip to Japan next month. Interested in vegetarian restaurants in Tokyo.",
-   "user_id":"alex123",
+   "user_id":"alex",
    "hash":"d2088c936e259f2f5d2d75543d31401c",
    "metadata":"None",
    "immutable": false,
@@ -1152,55 +1156,55 @@ You can filter memories by their categories when using get_all:
 
 ```python Python
 # Get memories with specific categories
-memories = client.get_all(user_id="alex123", categories=["likes"])
+memories = client.get_all(user_id="alex", categories=["likes"])
 
 # Get memories with multiple categories
-memories = client.get_all(user_id="alex123", categories=["likes", "food_preferences"])
+memories = client.get_all(user_id="alex", categories=["likes", "food_preferences"])
 
 # Custom pagination with categories
-memories = client.get_all(user_id="alex123", categories=["likes"], page=1, page_size=50)
+memories = client.get_all(user_id="alex", categories=["likes"], page=1, page_size=50)
 
 # Get memories with specific keywords
-memories = client.get_all(user_id="alex123", keywords="to play", page=1, page_size=50)
+memories = client.get_all(user_id="alex", keywords="to play", page=1, page_size=50)
 ```
 
 ```javascript JavaScript
 // Get memories with specific categories
-client.getAll({ user_id: "alex123", categories: ["likes"] })
+client.getAll({ user_id: "alex", categories: ["likes"] })
     .then(memories => console.log(memories))
     .catch(error => console.error(error));
 
 // Get memories with multiple categories
-client.getAll({ user_id: "alex123", categories: ["likes", "food_preferences"] })
+client.getAll({ user_id: "alex", categories: ["likes", "food_preferences"] })
     .then(memories => console.log(memories))
     .catch(error => console.error(error));
 
 // Custom pagination with categories
-client.getAll({ user_id: "alex123", categories: ["likes"], page: 1, page_size: 50 })
+client.getAll({ user_id: "alex", categories: ["likes"], page: 1, page_size: 50 })
     .then(memories => console.log(memories))
     .catch(error => console.error(error));
 
 // Get memories with specific keywords
-client.getAll({ user_id: "alex123", keywords: "to play", page: 1, page_size: 50 })
+client.getAll({ user_id: "alex", keywords: "to play", page: 1, page_size: 50 })
     .then(memories => console.log(memories))
     .catch(error => console.error(error));
 ```
 
 ```bash cURL
 # Get memories with specific categories
-curl -X GET "https://api.mem0.ai/v1/memories/?user_id=alex123&categories=likes" \
+curl -X GET "https://api.mem0.ai/v1/memories/?user_id=alex&categories=likes" \
      -H "Authorization: Token your-api-key"
 
 # Get memories with multiple categories
-curl -X GET "https://api.mem0.ai/v1/memories/?user_id=alex123&categories=likes,food_preferences" \
+curl -X GET "https://api.mem0.ai/v1/memories/?user_id=alex&categories=likes,food_preferences" \
      -H "Authorization: Token your-api-key"
 
 # Custom pagination with categories
-curl -X GET "https://api.mem0.ai/v1/memories/?user_id=alex123&categories=likes&page=1&page_size=50" \
+curl -X GET "https://api.mem0.ai/v1/memories/?user_id=alex&categories=likes&page=1&page_size=50" \
      -H "Authorization: Token your-api-key"
 
 # Get memories with specific keywords
-curl -X GET "https://api.mem0.ai/v1/memories/?user_id=alex123&keywords=to play&page=1&page_size=50" \
+curl -X GET "https://api.mem0.ai/v1/memories/?user_id=alex&keywords=to play&page=1&page_size=50" \
      -H "Authorization: Token your-api-key"
 ```
 
@@ -1213,7 +1217,7 @@ curl -X GET "https://api.mem0.ai/v1/memories/?user_id=alex123&keywords=to play&p
         {
             "id": "06d8df63-7bd2-4fad-9acb-60871bcecee0",
             "memory": "Likes pizza and pasta",
-            "user_id": "alex123",
+            "user_id": "alex",
             "hash": "d2088c936e259f2f5d2d75543d31401c",
             "metadata": null,
             "immutable": false,
@@ -1224,7 +1228,7 @@ curl -X GET "https://api.mem0.ai/v1/memories/?user_id=alex123&keywords=to play&p
         {
             "id": "b4229775-d860-4ccb-983f-0f628ca112f5",
             "memory": "Likes to travel to beach destinations",
-            "user_id": "alex123",
+            "user_id": "alex",
             "hash": "d2088c936e259f2f5d2d75543d31401c",
             "metadata": null,
             "immutable": false,
@@ -1585,7 +1589,7 @@ curl -X GET "https://api.mem0.ai/v1/memories/<memory-id-here>/history/" \
       ],
       "old_memory":"None",
       "new_memory":"Turned vegetarian.",
-      "user_id":"alex123456",
+      "user_id":"alex",
       "event":"ADD",
       "metadata":"None",
       "created_at":"2024-07-26T01:02:41.737310-07:00",

examples/mem0-demo/app/assistant.tsx

@@ -6,7 +6,7 @@ import { Thread } from "@/components/assistant-ui/thread";
 import { ThreadList } from "@/components/assistant-ui/thread-list";
 import { useEffect, useState } from "react";
 import { v4 as uuidv4 } from "uuid";
-import { Sun, Moon, MessageSquare } from "lucide-react";
+import { Sun, Moon, AlignJustify } from "lucide-react";
 import { Button } from "@/components/ui/button";
 import ThemeAwareLogo from "@/components/mem0/theme-aware-logo";
 import Link from "next/link";
@@ -63,7 +63,7 @@ export const Assistant = () => {
 
   return (
     <AssistantRuntimeProvider runtime={runtime}>
-      <div className={`h-dvh bg-[#f8fafc] dark:bg-zinc-900 text-[#1e293b] ${isDarkMode ? "dark" : ""}`}>
+      <div className={`bg-[#f8fafc] dark:bg-zinc-900 text-[#1e293b] ${isDarkMode ? "dark" : ""}`}>
         <header className="h-16 border-b border-[#e2e8f0] flex items-center justify-between px-4 sm:px-6 bg-white dark:bg-zinc-900 dark:border-zinc-800 dark:text-white">
           <div className="flex items-center">
           <Link href="/" className="flex items-center">
@@ -71,29 +71,34 @@ export const Assistant = () => {
           </Link>
           </div>
 
-          <div className="flex items-center">
-            <Button 
+          <Button 
               variant="ghost" 
               size="sm" 
               onClick={() => setSidebarOpen(true)}
               className="text-[#475569] dark:text-zinc-300 md:hidden"
             >
-              <MessageSquare className="w-10 h-10" />
-            </Button>
+              <AlignJustify size={24} className="md:hidden" />
+          </Button>
+
+
+          <div className="md:flex items-center hidden">
             <button
               className="p-2 rounded-full hover:bg-[#eef2ff] dark:hover:bg-zinc-800 text-[#475569] dark:text-zinc-300"
               onClick={toggleDarkMode}
               aria-label="Toggle theme"
             >
-              {isDarkMode ? <Sun className="w-5 h-5" /> : <Moon className="w-5 h-5" />}
+              {isDarkMode ? <Sun className="w-6 h-6" /> : <Moon className="w-6 h-6" />}
             </button>
             <GithubButton url="https://github.com/mem0ai/mem0/tree/main/examples" />
+
+            <Link href={"https://app.mem0.ai/"} target="_blank" className="py-2 ml-2 px-4 font-semibold dark:bg-zinc-100 dark:hover:bg-zinc-200 bg-zinc-800 text-white rounded-full hover:bg-zinc-900 dark:text-[#475569]">
+              Save Memories
+            </Link>
           </div>
         </header>
-
-        <div className="grid grid-cols-1 md:grid-cols-[260px_1fr] gap-x-0 h-[calc(100vh-8rem)] md:h-[calc(100vh-4rem)]">
+        <div className="grid grid-cols-1 md:grid-cols-[260px_1fr] gap-x-0 h-[calc(100dvh-4rem)]">
           <ThreadList onResetUserId={resetUserId} isDarkMode={isDarkMode} />
-          <Thread sidebarOpen={sidebarOpen} setSidebarOpen={setSidebarOpen} onResetUserId={resetUserId} isDarkMode={isDarkMode} />
+          <Thread sidebarOpen={sidebarOpen} setSidebarOpen={setSidebarOpen} onResetUserId={resetUserId} isDarkMode={isDarkMode} toggleDarkMode={toggleDarkMode} />
         </div>
       </div>
     </AssistantRuntimeProvider>

examples/mem0-demo/components/assistant-ui/thread-list.tsx

@@ -19,14 +19,14 @@ import {
   AlertDialogTitle,
   AlertDialogTrigger,
 } from "@/components/ui/alert-dialog";
-import ThemeAwareLogo from "@/components/assistant-ui/theme-aware-logo";
-import Link from "next/link";
+// import ThemeAwareLogo from "@/components/assistant-ui/theme-aware-logo";
+// import Link from "next/link";
 interface ThreadListProps {
   onResetUserId?: () => void;
   isDarkMode: boolean;
 }
 
-export const ThreadList: FC<ThreadListProps> = ({ onResetUserId, isDarkMode }) => {
+export const ThreadList: FC<ThreadListProps> = ({ onResetUserId }) => {
   const [open, setOpen] = useState(false);
   
   return (
@@ -79,13 +79,7 @@ export const ThreadList: FC<ThreadListProps> = ({ onResetUserId, isDarkMode }) =
           </div>
           <ThreadListItems />
         </div>
-        <div>
-          <Link href="https://www.assistant-ui.com/" target="_blank" className="flex justify-center items-center gap-2">
-          <h1 className="text-sm text-[#475569] dark:text-zinc-300 text-center">built using</h1>
-            <ThemeAwareLogo width={24} height={24} isDarkMode={isDarkMode} />
-            <p className="text-md font-bold dark:text-zinc-300">assistant-ui</p>
-          </Link>
-        </div>
+
       </ThreadListPrimitive.Root>
     </div>
   );

examples/mem0-demo/components/assistant-ui/thread.tsx

@@ -22,9 +22,12 @@ import {
   SendHorizontalIcon,
   ArchiveIcon,
   PlusIcon,
+  Sun,
+  Moon,
+  SaveIcon,
 } from "lucide-react";
 import { cn } from "@/lib/utils";
-import { Dispatch, SetStateAction, useState } from "react";
+import { Dispatch, SetStateAction, useState, useRef } from "react";
 import { Button } from "@/components/ui/button";
 import { ScrollArea } from "../ui/scroll-area";
 import { TooltipIconButton } from "@/components/assistant-ui/tooltip-icon-button";
@@ -42,14 +45,14 @@ import {
   AlertDialogTitle,
   AlertDialogTrigger,
 } from "@/components/ui/alert-dialog";
+import GithubButton from "../mem0/github-button";
 import Link from "next/link";
-import ThemeAwareLogo from "./theme-aware-logo";
-
 interface ThreadProps {
   sidebarOpen: boolean;
   setSidebarOpen: Dispatch<SetStateAction<boolean>>;
   onResetUserId?: () => void;
   isDarkMode: boolean;
+  toggleDarkMode: () => void;
 }
 
 export const Thread: FC<ThreadProps> = ({
@@ -57,12 +60,14 @@ export const Thread: FC<ThreadProps> = ({
   setSidebarOpen,
   onResetUserId,
   isDarkMode,
+  toggleDarkMode
 }) => {
   const [resetDialogOpen, setResetDialogOpen] = useState(false);
+  const composerInputRef = useRef<HTMLTextAreaElement>(null);
 
   return (
     <ThreadPrimitive.Root
-      className="bg-[#f8fafc] dark:bg-zinc-900 box-border h-full flex flex-col overflow-hidden relative"
+      className="bg-[#f8fafc] dark:bg-zinc-900 box-border flex flex-col overflow-hidden relative h-[calc(100dvh-4rem)] pb-4 md:h-full"
       style={{
         ["--thread-max-width" as string]: "42rem",
       }}
@@ -78,13 +83,13 @@ export const Thread: FC<ThreadProps> = ({
       {/* Mobile sidebar drawer */}
       <div
         className={cn(
-          "fixed inset-y-0 left-0 z-40 w-[85%] bg-white dark:bg-zinc-900 transform transition-transform duration-300 ease-in-out md:hidden",
+          "fixed inset-y-0 left-0 z-40 w-[75%] bg-white shadow-lg rounded-r-lg dark:bg-zinc-900 transform transition-transform duration-300 ease-in-out md:hidden",
           sidebarOpen ? "translate-x-0" : "-translate-x-full"
         )}
       >
         <div className="h-full flex flex-col">
           <div className="flex items-center justify-between border-b dark:text-white border-[#e2e8f0] dark:border-zinc-800 p-4">
-            <h2 className="font-medium">Recent Chats</h2>
+            <h2 className="font-medium">Settings</h2>
             <div className="flex items-center gap-2">
               {onResetUserId && (
                 <AlertDialog
@@ -141,13 +146,44 @@ export const Thread: FC<ThreadProps> = ({
             <div className="flex flex-col justify-between items-stretch gap-1.5 h-full dark:text-white">
               <ThreadListPrimitive.Root className="flex flex-col items-stretch gap-1.5 h-full dark:text-white">
                 <ThreadListPrimitive.New asChild>
+                  <div className="flex items-center flex-col gap-2 w-full">
                   <Button
-                    className="hover:bg-[#eef2ff] dark:hover:bg-zinc-800 dark:data-[active]:bg-zinc-800 flex items-center justify-start gap-1 rounded-lg px-2.5 py-2 text-start bg-[#4f46e5] text-white dark:bg-[#6366f1]"
+                    className="hover:bg-zinc-600 w-full dark:hover:bg-zinc-800 dark:data-[active]:bg-zinc-800 flex items-center justify-start gap-1 rounded-lg px-2.5 py-2 text-start bg-[#4f46e5] text-white dark:bg-[#6366f1]"
                     variant="default"
                   >
                     <PlusIcon className="w-4 h-4" />
                     New Thread
                   </Button>
+                    <Button
+                      className="hover:bg-zinc-600 w-full dark:hover:bg-zinc-700 dark:data-[active]:bg-zinc-800 flex items-center justify-start gap-1 rounded-lg px-2.5 py-2 text-start bg-zinc-800 text-white"
+                      onClick={toggleDarkMode}
+                      aria-label="Toggle theme"
+                    >
+                      {isDarkMode ? (
+                        <div className="flex items-center gap-2">
+                          <Sun className="w-6 h-6" /> 
+                          <span>Toggle Light Mode</span>
+                        </div>
+                      ) : (
+                        <div className="flex items-center gap-2">
+                          <Moon className="w-6 h-6" />
+                          <span>Toggle Dark Mode</span>
+                        </div>
+                      )}
+                    </Button>
+                    <GithubButton url="https://github.com/mem0ai/mem0/tree/main/examples" className="w-full rounded-lg h-9 pl-2 text-sm font-semibold bg-zinc-800 dark:border-zinc-800 dark:text-white text-white hover:bg-zinc-900" text="View on Github" />
+
+                    <Link
+                      href={"https://app.mem0.ai/"}
+                      target="_blank"
+                      className="py-2 px-4 w-full rounded-lg h-9 pl-3 text-sm font-semibold dark:bg-zinc-800 dark:hover:bg-zinc-700 bg-zinc-800 text-white hover:bg-zinc-900 dark:text-white"
+                    >
+                      <span className="flex items-center gap-2">
+                        <SaveIcon className="w-4 h-4" />
+                        Save Memories
+                      </span>
+                    </Link>
+                  </div>
                 </ThreadListPrimitive.New>
                 <div className="mt-4 mb-2">
                   <h2 className="text-sm font-medium text-[#475569] dark:text-zinc-300 px-2.5">
@@ -156,29 +192,18 @@ export const Thread: FC<ThreadProps> = ({
                 </div>
                 <ThreadListPrimitive.Items components={{ ThreadListItem }} />
               </ThreadListPrimitive.Root>
-              <div>
-                <Link
-                  href="https://www.assistant-ui.com/"
-                  target="_blank"
-                  className="flex justify-center items-center gap-2"
-                >
-                  <h1 className="text-sm text-[#475569] dark:text-zinc-300 text-center">
-                    built using
-                  </h1>
-                  <ThemeAwareLogo width={24} height={24} isDarkMode={isDarkMode} />
-                  <p className="text-md font-bold dark:text-zinc-300">
-                    assistant-ui
-                  </p>
-                </Link>
-              </div>
             </div>
           </div>
         </div>
       </div>
 
-      <ScrollArea className="flex-1">
-        <div className="flex h-full flex-col items-center px-4 pt-8 justify-end">
-          <ThreadWelcome />
+      <ScrollArea className="flex-1 w-full">
+        <div className="flex h-full flex-col w-full items-center px-4 pt-8 justify-end">
+          <ThreadWelcome
+            composerInputRef={
+              composerInputRef as React.RefObject<HTMLTextAreaElement>
+            }
+          />
 
           <ThreadPrimitive.Messages
             components={{
@@ -194,9 +219,13 @@ export const Thread: FC<ThreadProps> = ({
         </div>
       </ScrollArea>
 
-      <div className="sticky bottom-0 mt-3 flex w-full max-w-[var(--thread-max-width)] flex-col items-center justify-end rounded-t-lg bg-inherit px-4 pb-4 mx-auto">
+      <div className="sticky bottom-0 flex w-full max-w-[var(--thread-max-width)] flex-col items-center justify-end rounded-t-lg bg-inherit px-4 md:pb-4 mx-auto">
         <ThreadScrollToBottom />
-        <Composer />
+        <Composer
+          composerInputRef={
+            composerInputRef as React.RefObject<HTMLTextAreaElement>
+          }
+        />
       </div>
     </ThreadPrimitive.Root>
   );
@@ -216,57 +245,74 @@ const ThreadScrollToBottom: FC = () => {
   );
 };
 
-const ThreadWelcome: FC = () => {
+interface ThreadWelcomeProps {
+  composerInputRef: React.RefObject<HTMLTextAreaElement>;
+}
+
+const ThreadWelcome: FC<ThreadWelcomeProps> = ({ composerInputRef }) => {
   return (
     <ThreadPrimitive.Empty>
-      <div className="flex w-full max-w-[var(--thread-max-width)] flex-grow flex-col">
-        <div className="flex w-full flex-grow flex-col items-center justify-start h-[calc(100vh-23rem)] md:h-[calc(100vh-18rem)]">
+      <div className="flex w-full flex-grow flex-col mt-8 md:h-[calc(100vh-15rem)]">
+        <div className="flex w-full flex-grow flex-col items-center justify-start">
           <div className="flex flex-col items-center justify-center h-full">
-            <div className="text-2xl md:text-4xl font-bold text-[#1e293b] dark:text-white mb-2">
+            <div className="text-[2rem] leading-[1] tracking-[-0.02em] md:text-4xl font-bold text-[#1e293b] dark:text-white mb-2 text-center md:w-full w-5/6">
               Mem0 - ChatGPT with memory
             </div>
-            <p className="text-center text-sm text-[#1e293b] dark:text-white mb-2 w-3/4">
+            <p className="text-center text-md text-[#1e293b] dark:text-white mb-2 md:w-3/4 w-5/6">
               A personalized AI chat app powered by Mem0 that remembers your
               preferences, facts, and memories.
             </p>
           </div>
         </div>
-        <div className="flex flex-col items-center justify-center">
+        <div className="flex flex-col items-center justify-center mt-16">
           <p className="mt-4 font-medium text-[#1e293b] dark:text-white">
             How can I help you today?
           </p>
-          <ThreadWelcomeSuggestions />
+          <ThreadWelcomeSuggestions composerInputRef={composerInputRef} />
         </div>
       </div>
     </ThreadPrimitive.Empty>
   );
 };
 
-const ThreadWelcomeSuggestions: FC = () => {
+interface ThreadWelcomeSuggestionsProps {
+  composerInputRef: React.RefObject<HTMLTextAreaElement>;
+}
+
+const ThreadWelcomeSuggestions: FC<ThreadWelcomeSuggestionsProps> = ({ composerInputRef }) => {
   return (
-    <div className="mt-3 flex w-full items-stretch justify-center gap-4 dark:text-white">
+    <div className="mt-3 flex flex-col md:flex-row w-full md:items-stretch justify-center gap-4 dark:text-white items-center">
       <ThreadPrimitive.Suggestion
-        className="hover:bg-[#eef2ff] dark:hover:bg-zinc-800 flex max-w-sm grow basis-0 flex-col items-center justify-center rounded-[2rem] border border-[#e2e8f0] dark:border-zinc-700 p-3 transition-colors ease-in"
+        className="hover:bg-[#eef2ff] w-full dark:hover:bg-zinc-800 flex max-w-sm grow basis-0 flex-col items-center justify-center rounded-[2rem] border border-[#e2e8f0] dark:border-zinc-700 p-3 transition-colors ease-in"
         prompt="I like to travel to "
         method="replace"
+        onClick={() => {
+          composerInputRef.current?.focus();
+        }}
       >
         <span className="line-clamp-2 text-ellipsis text-sm font-semibold">
           Travel
         </span>
       </ThreadPrimitive.Suggestion>
       <ThreadPrimitive.Suggestion
-        className="hover:bg-[#eef2ff] dark:hover:bg-zinc-800 flex max-w-sm grow basis-0 flex-col items-center justify-center rounded-[2rem] border border-[#e2e8f0] dark:border-zinc-700 p-3 transition-colors ease-in"
+        className="hover:bg-[#eef2ff] w-full dark:hover:bg-zinc-800 flex max-w-sm grow basis-0 flex-col items-center justify-center rounded-[2rem] border border-[#e2e8f0] dark:border-zinc-700 p-3 transition-colors ease-in"
         prompt="I like to eat "
         method="replace"
+        onClick={() => {
+          composerInputRef.current?.focus();
+        }}
       >
         <span className="line-clamp-2 text-ellipsis text-sm font-semibold">
           Food
         </span>
       </ThreadPrimitive.Suggestion>
       <ThreadPrimitive.Suggestion
-        className="hover:bg-[#eef2ff] dark:hover:bg-zinc-800 flex max-w-sm grow basis-0 flex-col items-center justify-center rounded-[2rem] border border-[#e2e8f0] dark:border-zinc-700 p-3 transition-colors ease-in"
+        className="hover:bg-[#eef2ff] w-full dark:hover:bg-zinc-800 flex max-w-sm grow basis-0 flex-col items-center justify-center rounded-[2rem] border border-[#e2e8f0] dark:border-zinc-700 p-3 transition-colors ease-in"
         prompt="I am working on "
         method="replace"
+        onClick={() => {
+          composerInputRef.current?.focus();
+        }}
       >
         <span className="line-clamp-2 text-ellipsis text-sm font-semibold">
           Project details
@@ -276,7 +322,11 @@ const ThreadWelcomeSuggestions: FC = () => {
   );
 };
 
-const Composer: FC = () => {
+interface ComposerProps {
+  composerInputRef: React.RefObject<HTMLTextAreaElement>;
+}
+
+const Composer: FC<ComposerProps> = ({ composerInputRef }) => {
   return (
     <ComposerPrimitive.Root className="focus-within:border-[#4f46e5]/20 dark:focus-within:border-[#6366f1]/20 flex w-full flex-wrap items-end rounded-full border border-[#e2e8f0] dark:border-zinc-700 bg-white dark:bg-zinc-800 px-2.5 shadow-sm transition-colors ease-in">
       <ComposerPrimitive.Input
@@ -284,6 +334,7 @@ const Composer: FC = () => {
         autoFocus
         placeholder="Message to Mem0..."
         className="placeholder:text-zinc-400 dark:placeholder:text-zinc-500 max-h-40 flex-grow resize-none border-none bg-transparent px-2 py-4 text-sm outline-none focus:ring-0 disabled:cursor-not-allowed text-[#1e293b] dark:text-zinc-200"
+        ref={composerInputRef}
       />
       <ComposerAction />
     </ComposerPrimitive.Root>

examples/mem0-demo/components/mem0/github-button.tsx

@@ -1,18 +1,18 @@
+import { cn } from "@/lib/utils";
 
-
-const GithubButton = ({ url }: { url: string }) => {
+const GithubButton = ({ url, className, text }: { url: string, className?: string, text?: string }) => {
   return (
     <a
       href={url}
       target="_blank"
       rel="noopener noreferrer"
-      className="flex items-center bg-black text-white rounded-full shadow-lg hover:bg-gray-800 transition border border-gray-700"
+      className={cn("flex items-center bg-black text-white rounded-full shadow-lg hover:bg-gray-800 transition border border-gray-700", className)}
     >
       <svg
         xmlns="http://www.w3.org/2000/svg"
         viewBox="0 0 24 24"
         fill="white"
-        className="w-6 h-6"
+        className="w-5 h-5 md:w-6 md:h-6"
       >
         <path
           fillRule="evenodd"
@@ -20,6 +20,7 @@ const GithubButton = ({ url }: { url: string }) => {
           clipRule="evenodd"
         />
       </svg>
+      {text && <span className="ml-2">{text}</span>}
     </a>
   );
 };

examples/openai-inbuilt-tools/index.js

@@ -0,0 +1,83 @@
+import MemoryClient from "mem0ai";
+import { OpenAI } from "openai";
+import { zodResponsesFunction } from "openai/helpers/zod";
+import { z } from "zod";
+
+const mem0Config = {
+    apiKey: process.env.MEM0_API_KEY, // GET THIS API KEY FROM MEM0 (https://app.mem0.ai/dashboard/api-keys)
+    user_id: "sample-user",
+};
+
+async function run() {
+    // RESPONES WITHOUT MEMORIES
+    console.log("\n\nRESPONES WITHOUT MEMORIES\n\n");
+    await main();
+
+    // ADDING SOME SAMPLE MEMORIES
+    await addSampleMemories();
+
+    // RESPONES WITH MEMORIES
+    console.log("\n\nRESPONES WITH MEMORIES\n\n");
+    await main(true);
+}
+
+// OpenAI Response Schema
+const CarSchema = z.object({
+  car_name: z.string(),
+  car_price: z.string(),
+  car_url: z.string(),
+  car_image: z.string(),
+  car_description: z.string(),
+});
+
+const Cars = z.object({
+  cars: z.array(CarSchema),
+});
+
+async function main(memory = false) {
+  const openAIClient = new OpenAI();
+  const mem0Client = new MemoryClient(mem0Config);
+
+  const input = "Suggest me some cars that I can buy today.";
+
+  const tool = zodResponsesFunction({ name: "carRecommendations", parameters: Cars });
+
+  // First, let's store the user's memories from user input if any
+  await mem0Client.add([{
+    role: "user",
+    content: input,
+  }], mem0Config);
+
+  // Then search for relevant memories
+  let relevantMemories = []
+  if (memory) {
+    relevantMemories = await mem0Client.search(input, mem0Config);
+  }
+
+  const response = await openAIClient.responses.create({
+    model: "gpt-4o",
+    tools: [{ type: "web_search_preview" }, tool],
+    input: `${getMemoryString(relevantMemories)}\n${input}`,
+  });
+
+  console.log(response.output);
+}
+
+async function addSampleMemories() {
+  const mem0Client = new MemoryClient(mem0Config);
+
+  const myInterests = "I Love BMW, Audi and Porsche. I Hate Mercedes. I love Red cars and Maroon cars. I have a budget of 120K to 150K USD. I like Audi the most.";
+  
+  await mem0Client.add([{
+    role: "user",
+    content: myInterests,
+  }], mem0Config);
+}
+
+const getMemoryString = (memories) => {
+    const MEMORY_STRING_PREFIX = "These are the memories I have stored. Give more weightage to the question by users and try to answer that first. You have to modify your answer based on the memories I have provided. If the memories are irrelevant you can ignore them. Also don't reply to this section of the prompt, or the memories, they are only for your reference. The MEMORIES of the USER are: \n\n";
+    const memoryString = memories.map((mem) => `${mem.memory}`).join("\n") ?? "";
+    return memoryString.length > 0 ? `${MEMORY_STRING_PREFIX}${memoryString}` : "";
+};
+
+run().catch(console.error);

examples/openai-inbuilt-tools/package.json

@@ -0,0 +1,19 @@
+{
+  "name": "openai-inbuilt-tools",
+  "version": "1.0.0",
+  "description": "",
+  "license": "ISC",
+  "author": "",
+  "type": "module",
+  "main": "index.js",
+  "scripts": {
+    "test": "echo \"Error: no test specified\" && exit 1",
+    "start": "node index.js"
+  },
+  "packageManager": "pnpm@10.5.2+sha512.da9dc28cd3ff40d0592188235ab25d3202add8a207afbedc682220e4a0029ffbff4562102b9e6e46b4e3f9e8bd53e6d05de48544b0c57d4b0179e22c76d1199b",
+  "dependencies": {
+    "mem0ai": "^2.1.2",
+    "openai": "^4.87.2",
+    "zod": "^3.24.2"
+  }
+}

mem0-ts/package.json

@@ -1,6 +1,6 @@
 {
   "name": "mem0ai",
-  "version": "2.1.1",
+  "version": "2.1.5",
   "description": "The Memory Layer For Your AI Apps",
   "main": "./dist/index.js",
   "module": "./dist/index.mjs",

mem0-ts/src/client/mem0.ts

@@ -205,6 +205,10 @@ export default class MemoryClient {
       if (options.project_name) delete options.project_name;
     }
 
+    if (options.api_version) {
+      options.version = options.api_version.toString();
+    }
+
     const payload = this._preparePayload(messages, options);
     const response = await this._fetchWithErrorHandling(
       `${this.host}/v1/memories/`,

mem0-ts/src/client/mem0.types.ts

@@ -1,4 +1,6 @@
 export interface MemoryOptions {
+  api_version?: API_VERSION | string;
+  version?: API_VERSION | string;
   user_id?: string;
   agent_id?: string;
   app_id?: string;
@@ -17,6 +19,7 @@ export interface MemoryOptions {
   enable_graph?: boolean;
   start_date?: string;
   end_date?: string;
+  custom_categories?: custom_categories[];
 }
 
 export interface ProjectOptions {

mem0-ts/src/client/telemetry.browser.ts

@@ -4,7 +4,7 @@ import type { TelemetryClient } from "./telemetry.types";
 
 let version = "1.0.20";
 
-const MEM0_TELEMETRY = process.env.MEM0_TELEMETRY !== "false";
+const MEM0_TELEMETRY = "false";
 const POSTHOG_API_KEY = "phc_hgJkUVJFYtmaJqrvf6CYN67TIQ8yhXAkWzUn9AMU4yX";
 const POSTHOG_HOST = "https://us.i.posthog.com";
 

mem0-ts/src/oss/src/vector_stores/redis.ts

@@ -62,7 +62,7 @@ interface RedisDocument {
     run_id?: string;
     user_id?: string;
     metadata?: string;
-    vector_score?: number;
+    __vector_score?: number;
   };
 }
 
@@ -108,6 +108,30 @@ const EXCLUDED_KEYS = new Set([
   "updated_at",
 ]);
 
+// Utility function to convert object keys to snake_case
+function toSnakeCase(obj: Record<string, any>): Record<string, any> {
+  if (typeof obj !== "object" || obj === null) return obj;
+
+  return Object.fromEntries(
+    Object.entries(obj).map(([key, value]) => [
+      key.replace(/[A-Z]/g, (letter) => `_${letter.toLowerCase()}`),
+      value,
+    ]),
+  );
+}
+
+// Utility function to convert object keys to camelCase
+function toCamelCase(obj: Record<string, any>): Record<string, any> {
+  if (typeof obj !== "object" || obj === null) return obj;
+
+  return Object.fromEntries(
+    Object.entries(obj).map(([key, value]) => [
+      key.replace(/_([a-z])/g, (_, letter) => letter.toUpperCase()),
+      value,
+    ]),
+  );
+}
+
 export class RedisDB implements VectorStore {
   private client: RedisClientType<
     RedisDefaultModules & RedisModules & RedisFunctions & RedisScripts
@@ -272,7 +296,7 @@ export class RedisDB implements VectorStore {
     payloads: Record<string, any>[],
   ): Promise<void> {
     const data = vectors.map((vector, idx) => {
-      const payload = payloads[idx];
+      const payload = toSnakeCase(payloads[idx]);
       const id = ids[idx];
 
       // Create entry with required fields
@@ -322,8 +346,9 @@ export class RedisDB implements VectorStore {
     limit: number = 5,
     filters?: SearchFilters,
   ): Promise<VectorStoreResult[]> {
-    const filterExpr = filters
-      ? Object.entries(filters)
+    const snakeFilters = filters ? toSnakeCase(filters) : undefined;
+    const filterExpr = snakeFilters
+      ? Object.entries(snakeFilters)
           .filter(([_, value]) => value !== null)
           .map(([key, value]) => `@${key}:{${value}}`)
           .join(" ")
@@ -344,8 +369,9 @@ export class RedisDB implements VectorStore {
         "memory",
         "metadata",
         "created_at",
+        "__vector_score",
       ],
-      SORTBY: "vector_score",
+      SORTBY: "__vector_score",
       DIALECT: 2,
       LIMIT: {
         from: 0,
@@ -356,12 +382,12 @@ export class RedisDB implements VectorStore {
     try {
       const results = (await this.client.ft.search(
         this.indexName,
-        `${filterExpr} =>[KNN ${limit} @embedding $vec AS vector_score]`,
+        `${filterExpr} =>[KNN ${limit} @embedding $vec AS __vector_score]`,
         searchOptions,
       )) as unknown as RedisSearchResult;
 
       return results.documents.map((doc) => {
-        const payload = {
+        const resultPayload = {
           hash: doc.value.hash,
           data: doc.value.memory,
           created_at: new Date(parseInt(doc.value.created_at)).toISOString(),
@@ -376,8 +402,8 @@ export class RedisDB implements VectorStore {
 
         return {
           id: doc.value.memory_id,
-          payload,
-          score: doc.value.vector_score,
+          payload: toCamelCase(resultPayload),
+          score: Number(doc.value.__vector_score) ?? 0,
         };
       });
     } catch (error) {
@@ -493,26 +519,27 @@ export class RedisDB implements VectorStore {
     vector: number[],
     payload: Record<string, any>,
   ): Promise<void> {
+    const snakePayload = toSnakeCase(payload);
     const entry: Record<string, any> = {
       memory_id: vectorId,
-      hash: payload.hash,
-      memory: payload.data,
-      created_at: new Date(payload.created_at).getTime(),
-      updated_at: new Date(payload.updated_at).getTime(),
+      hash: snakePayload.hash,
+      memory: snakePayload.data,
+      created_at: new Date(snakePayload.created_at).getTime(),
+      updated_at: new Date(snakePayload.updated_at).getTime(),
       embedding: Buffer.from(new Float32Array(vector).buffer),
     };
 
     // Add optional fields
     ["agent_id", "run_id", "user_id"].forEach((field) => {
-      if (field in payload) {
-        entry[field] = payload[field];
+      if (field in snakePayload) {
+        entry[field] = snakePayload[field];
       }
     });
 
     // Add metadata excluding specific keys
     entry.metadata = JSON.stringify(
       Object.fromEntries(
-        Object.entries(payload).filter(([key]) => !EXCLUDED_KEYS.has(key)),
+        Object.entries(snakePayload).filter(([key]) => !EXCLUDED_KEYS.has(key)),
       ),
     );
 
@@ -557,8 +584,9 @@ export class RedisDB implements VectorStore {
     filters?: SearchFilters,
     limit: number = 100,
   ): Promise<[VectorStoreResult[], number]> {
-    const filterExpr = filters
-      ? Object.entries(filters)
+    const snakeFilters = filters ? toSnakeCase(filters) : undefined;
+    const filterExpr = snakeFilters
+      ? Object.entries(snakeFilters)
           .filter(([_, value]) => value !== null)
           .map(([key, value]) => `@${key}:{${value}}`)
           .join(" ")
@@ -581,7 +609,7 @@ export class RedisDB implements VectorStore {
 
     const items = results.documents.map((doc) => ({
       id: doc.value.memory_id,
-      payload: {
+      payload: toCamelCase({
         hash: doc.value.hash,
         data: doc.value.memory,
         created_at: new Date(parseInt(doc.value.created_at)).toISOString(),
@@ -592,7 +620,7 @@ export class RedisDB implements VectorStore {
         ...(doc.value.run_id && { run_id: doc.value.run_id }),
         ...(doc.value.user_id && { user_id: doc.value.user_id }),
         ...JSON.parse(doc.value.metadata || "{}"),
-      },
+      }),
     }));
 
     return [items, results.total];

mem0/client/main.py

@@ -6,7 +6,7 @@ from typing import Any, Dict, List, Optional, Union
 
 import httpx
 
-from mem0.memory.setup import setup_config, get_user_id
+from mem0.memory.setup import get_user_id, setup_config
 from mem0.memory.telemetry import capture_client_event
 
 logger = logging.getLogger(__name__)
@@ -110,7 +110,7 @@ class MemoryClient:
             try:
                 error_data = e.response.json()
                 error_message = error_data.get("detail", str(e))
-            except:
+            except Exception:
                 error_message = str(e)
             raise ValueError(f"Error: {error_message}")
 

mem0/configs/base.py

@@ -48,8 +48,12 @@ class MemoryConfig(BaseModel):
         description="The version of the API",
         default="v1.1",
     )
-    custom_prompt: Optional[str] = Field(
-        description="Custom prompt for the memory",
+    custom_fact_extraction_prompt: Optional[str] = Field(
+        description="Custom prompt for the fact extraction",
+        default=None,
+    )
+    custom_update_memory_prompt: Optional[str] = Field(
+        description="Custom prompt for the update memory",
         default=None,
     )
 

mem0/configs/prompts.py

@@ -58,169 +58,189 @@ Following is a conversation between the user and the assistant. You have to extr
 You should detect the language of the user input and record the facts in the same language.
 """
 
+DEFAULT_UPDATE_MEMORY_PROMPT = """You are a smart memory manager which controls the memory of a system.
+You can perform four operations: (1) add into the memory, (2) update the memory, (3) delete from the memory, and (4) no change.
 
-def get_update_memory_messages(retrieved_old_memory_dict, response_content):
-    return f"""You are a smart memory manager which controls the memory of a system.
-    You can perform four operations: (1) add into the memory, (2) update the memory, (3) delete from the memory, and (4) no change.
+Based on the above four operations, the memory will change.
 
-    Based on the above four operations, the memory will change.
+Compare newly retrieved facts with the existing memory. For each new fact, decide whether to:
+- ADD: Add it to the memory as a new element
+- UPDATE: Update an existing memory element
+- DELETE: Delete an existing memory element
+- NONE: Make no change (if the fact is already present or irrelevant)
 
-    Compare newly retrieved facts with the existing memory. For each new fact, decide whether to:
-    - ADD: Add it to the memory as a new element
-    - UPDATE: Update an existing memory element
-    - DELETE: Delete an existing memory element
-    - NONE: Make no change (if the fact is already present or irrelevant)
+There are specific guidelines to select which operation to perform:
 
-    There are specific guidelines to select which operation to perform:
+1. **Add**: If the retrieved facts contain new information not present in the memory, then you have to add it by generating a new ID in the id field.
+- **Example**:
+    - Old Memory:
+        [
+            {
+                "id" : "0",
+                "text" : "User is a software engineer"
+            }
+        ]
+    - Retrieved facts: ["Name is John"]
+    - New Memory:
+        {
+            "memory" : [
+                {
+                    "id" : "0",
+                    "text" : "User is a software engineer",
+                    "event" : "NONE"
+                },
+                {
+                    "id" : "1",
+                    "text" : "Name is John",
+                    "event" : "ADD"
+                }
+            ]
 
-    1. **Add**: If the retrieved facts contain new information not present in the memory, then you have to add it by generating a new ID in the id field.
-        - **Example**:
-            - Old Memory:
-                [
-                    {{
-                        "id" : "0",
-                        "text" : "User is a software engineer"
-                    }}
-                ]
-            - Retrieved facts: ["Name is John"]
-            - New Memory:
-                {{
-                    "memory" : [
-                        {{
-                            "id" : "0",
-                            "text" : "User is a software engineer",
-                            "event" : "NONE"
-                        }},
-                        {{
-                            "id" : "1",
-                            "text" : "Name is John",
-                            "event" : "ADD"
-                        }}
-                    ]
+        }
 
-                }}
-
-    2. **Update**: If the retrieved facts contain information that is already present in the memory but the information is totally different, then you have to update it. 
-        If the retrieved fact contains information that conveys the same thing as the elements present in the memory, then you have to keep the fact which has the most information. 
-        Example (a) -- if the memory contains "User likes to play cricket" and the retrieved fact is "Loves to play cricket with friends", then update the memory with the retrieved facts.
-        Example (b) -- if the memory contains "Likes cheese pizza" and the retrieved fact is "Loves cheese pizza", then you do not need to update it because they convey the same information.
-        If the direction is to update the memory, then you have to update it.
-        Please keep in mind while updating you have to keep the same ID.
-        Please note to return the IDs in the output from the input IDs only and do not generate any new ID.
-        - **Example**:
-            - Old Memory:
-                [
-                    {{
-                        "id" : "0",
-                        "text" : "I really like cheese pizza"
-                    }},
-                    {{
-                        "id" : "1",
-                        "text" : "User is a software engineer"
-                    }},
-                    {{
-                        "id" : "2",
-                        "text" : "User likes to play cricket"
-                    }}
-                ]
-            - Retrieved facts: ["Loves chicken pizza", "Loves to play cricket with friends"]
-            - New Memory:
-                {{
-                "memory" : [
-                        {{
-                            "id" : "0",
-                            "text" : "Loves cheese and chicken pizza",
-                            "event" : "UPDATE",
-                            "old_memory" : "I really like cheese pizza"
-                        }},
-                        {{
-                            "id" : "1",
-                            "text" : "User is a software engineer",
-                            "event" : "NONE"
-                        }},
-                        {{
-                            "id" : "2",
-                            "text" : "Loves to play cricket with friends",
-                            "event" : "UPDATE",
-                            "old_memory" : "User likes to play cricket"
-                        }}
-                    ]
-                }}
+2. **Update**: If the retrieved facts contain information that is already present in the memory but the information is totally different, then you have to update it. 
+If the retrieved fact contains information that conveys the same thing as the elements present in the memory, then you have to keep the fact which has the most information. 
+Example (a) -- if the memory contains "User likes to play cricket" and the retrieved fact is "Loves to play cricket with friends", then update the memory with the retrieved facts.
+Example (b) -- if the memory contains "Likes cheese pizza" and the retrieved fact is "Loves cheese pizza", then you do not need to update it because they convey the same information.
+If the direction is to update the memory, then you have to update it.
+Please keep in mind while updating you have to keep the same ID.
+Please note to return the IDs in the output from the input IDs only and do not generate any new ID.
+- **Example**:
+    - Old Memory:
+        [
+            {
+                "id" : "0",
+                "text" : "I really like cheese pizza"
+            },
+            {
+                "id" : "1",
+                "text" : "User is a software engineer"
+            },
+            {
+                "id" : "2",
+                "text" : "User likes to play cricket"
+            }
+        ]
+    - Retrieved facts: ["Loves chicken pizza", "Loves to play cricket with friends"]
+    - New Memory:
+        {
+        "memory" : [
+                {
+                    "id" : "0",
+                    "text" : "Loves cheese and chicken pizza",
+                    "event" : "UPDATE",
+                    "old_memory" : "I really like cheese pizza"
+                },
+                {
+                    "id" : "1",
+                    "text" : "User is a software engineer",
+                    "event" : "NONE"
+                },
+                {
+                    "id" : "2",
+                    "text" : "Loves to play cricket with friends",
+                    "event" : "UPDATE",
+                    "old_memory" : "User likes to play cricket"
+                }
+            ]
+        }
 
 
-    3. **Delete**: If the retrieved facts contain information that contradicts the information present in the memory, then you have to delete it. Or if the direction is to delete the memory, then you have to delete it.
-        Please note to return the IDs in the output from the input IDs only and do not generate any new ID.
-        - **Example**:
-            - Old Memory:
-                [
-                    {{
-                        "id" : "0",
-                        "text" : "Name is John"
-                    }},
-                    {{
-                        "id" : "1",
-                        "text" : "Loves cheese pizza"
-                    }}
-                ]
-            - Retrieved facts: ["Dislikes cheese pizza"]
-            - New Memory:
-                {{
-                "memory" : [
-                        {{
-                            "id" : "0",
-                            "text" : "Name is John",
-                            "event" : "NONE"
-                        }},
-                        {{
-                            "id" : "1",
-                            "text" : "Loves cheese pizza",
-                            "event" : "DELETE"
-                        }}
-                ]
-                }}
+3. **Delete**: If the retrieved facts contain information that contradicts the information present in the memory, then you have to delete it. Or if the direction is to delete the memory, then you have to delete it.
+Please note to return the IDs in the output from the input IDs only and do not generate any new ID.
+- **Example**:
+    - Old Memory:
+        [
+            {
+                "id" : "0",
+                "text" : "Name is John"
+            },
+            {
+                "id" : "1",
+                "text" : "Loves cheese pizza"
+            }
+        ]
+    - Retrieved facts: ["Dislikes cheese pizza"]
+    - New Memory:
+        {
+        "memory" : [
+                {
+                    "id" : "0",
+                    "text" : "Name is John",
+                    "event" : "NONE"
+                },
+                {
+                    "id" : "1",
+                    "text" : "Loves cheese pizza",
+                    "event" : "DELETE"
+                }
+        ]
+        }
 
-    4. **No Change**: If the retrieved facts contain information that is already present in the memory, then you do not need to make any changes.
-        - **Example**:
-            - Old Memory:
-                [
-                    {{
-                        "id" : "0",
-                        "text" : "Name is John"
-                    }},
-                    {{
-                        "id" : "1",
-                        "text" : "Loves cheese pizza"
-                    }}
-                ]
-            - Retrieved facts: ["Name is John"]
-            - New Memory:
-                {{
-                "memory" : [
-                        {{
-                            "id" : "0",
-                            "text" : "Name is John",
-                            "event" : "NONE"
-                        }},
-                        {{
-                            "id" : "1",
-                            "text" : "Loves cheese pizza",
-                            "event" : "NONE"
-                        }}
-                    ]
-                }}
+4. **No Change**: If the retrieved facts contain information that is already present in the memory, then you do not need to make any changes.
+- **Example**:
+    - Old Memory:
+        [
+            {
+                "id" : "0",
+                "text" : "Name is John"
+            },
+            {
+                "id" : "1",
+                "text" : "Loves cheese pizza"
+            }
+        ]
+    - Retrieved facts: ["Name is John"]
+    - New Memory:
+        {
+        "memory" : [
+                {
+                    "id" : "0",
+                    "text" : "Name is John",
+                    "event" : "NONE"
+                },
+                {
+                    "id" : "1",
+                    "text" : "Loves cheese pizza",
+                    "event" : "NONE"
+                }
+            ]
+        }
+"""
 
+def get_update_memory_messages(retrieved_old_memory_dict, response_content, custom_update_memory_prompt=None):
+    if custom_update_memory_prompt is None:
+        global DEFAULT_UPDATE_MEMORY_PROMPT
+        custom_update_memory_prompt = DEFAULT_UPDATE_MEMORY_PROMPT
+        
+    return f"""{custom_update_memory_prompt}
+    
     Below is the current content of my memory which I have collected till now. You have to update it in the following format only:
 
-    ``
+    ```
     {retrieved_old_memory_dict}
-    ``
+    ```
 
     The new retrieved facts are mentioned in the triple backticks. You have to analyze the new retrieved facts and determine whether these facts should be added, updated, or deleted in the memory.
 
     ```
     {response_content}
     ```
-
+    
+    You must return your response in the following JSON structure only:
+    
+    {{
+        "memory" : [
+            {{
+                "id" : "<ID of the memory>",                # Use existing ID for updates/deletes, or new ID for additions
+                "text" : "<Content of the memory>",         # Content of the memory
+                "event" : "<Operation to be performed>",    # Must be "ADD", "UPDATE", "DELETE", or "NONE"
+                "old_memory" : "<Old memory content>"       # Required only if the event is "UPDATE"
+            }},
+            ...
+        ]
+    }}
+    
     Follow the instruction mentioned below:
     - Do not return anything from the custom few shot prompts provided above.
     - If the current memory is empty, then you have to add the new retrieved facts to the memory.
@@ -230,4 +250,4 @@ def get_update_memory_messages(retrieved_old_memory_dict, response_content):
     - If there is an update, the ID key should remain the same and only the value needs to be updated.
 
     Do not return anything except the JSON format.
-    """
+    """

mem0/configs/vector_stores/azure_ai_search.py

@@ -1,27 +1,53 @@
-from typing import Any, Dict
-
+from typing import Any, Dict, Optional
 from pydantic import BaseModel, Field, model_validator
 
 
 class AzureAISearchConfig(BaseModel):
     collection_name: str = Field("mem0", description="Name of the collection")
-    service_name: str = Field(None, description="Azure Cognitive Search service name")
-    api_key: str = Field(None, description="API key for the Azure Cognitive Search service")
+    service_name: str = Field(None, description="Azure AI Search service name")
+    api_key: str = Field(None, description="API key for the Azure AI Search service")
     embedding_model_dims: int = Field(None, description="Dimension of the embedding vector")
-    use_compression: bool = Field(False, description="Whether to use scalar quantization vector compression.")
-
+    compression_type: Optional[str] = Field(
+        None, 
+        description="Type of vector compression to use. Options: 'scalar', 'binary', or None"
+    )
+    use_float16: bool = Field(
+        False, 
+        description="Whether to store vectors in half precision (Edm.Half) instead of full precision (Edm.Single)"
+    )
+    
     @model_validator(mode="before")
     @classmethod
     def validate_extra_fields(cls, values: Dict[str, Any]) -> Dict[str, Any]:
         allowed_fields = set(cls.model_fields.keys())
         input_fields = set(values.keys())
         extra_fields = input_fields - allowed_fields
+        
+        # Check for use_compression to provide a helpful error
+        if "use_compression" in extra_fields:
+            raise ValueError(
+                "The parameter 'use_compression' is no longer supported. "
+                "Please use 'compression_type=\"scalar\"' instead of 'use_compression=True' "
+                "or 'compression_type=None' instead of 'use_compression=False'."
+            )
+        
         if extra_fields:
             raise ValueError(
-                f"Extra fields not allowed: {', '.join(extra_fields)}. Please input only the following fields: {', '.join(allowed_fields)}"
+                f"Extra fields not allowed: {', '.join(extra_fields)}. "
+                f"Please input only the following fields: {', '.join(allowed_fields)}"
             )
+        
+        # Validate compression_type values
+        if "compression_type" in values and values["compression_type"] is not None:
+            valid_types = ["scalar", "binary"]
+            if values["compression_type"].lower() not in valid_types:
+                raise ValueError(
+                    f"Invalid compression_type: {values['compression_type']}. "
+                    f"Must be one of: {', '.join(valid_types)}, or None"
+                )
+            
         return values
-
+    
     model_config = {
         "arbitrary_types_allowed": True,
-    }
+    }

mem0/configs/vector_stores/elasticsearch.py

@@ -1,4 +1,5 @@
-from typing import Any, Dict, Optional
+from collections.abc import Callable
+from typing import Any, Dict, List, Optional
 
 from pydantic import BaseModel, Field, model_validator
 
@@ -15,6 +16,10 @@ class ElasticsearchConfig(BaseModel):
     verify_certs: bool = Field(True, description="Verify SSL certificates")
     use_ssl: bool = Field(True, description="Use SSL for connection")
     auto_create_index: bool = Field(True, description="Automatically create index during initialization")
+    custom_search_query: Optional[Callable[[List[float], int, Optional[Dict]], Dict]] = Field(
+        None,
+        description="Custom search query function. Parameters: (query, limit, filters) -> Dict"
+    )
 
     @model_validator(mode="before")
     @classmethod

mem0/configs/vector_stores/weaviate.py

@@ -0,0 +1,42 @@
+from typing import Any, ClassVar, Dict, Optional
+from pydantic import BaseModel, Field, model_validator
+
+
+class WeaviateConfig(BaseModel):
+    from weaviate import WeaviateClient
+
+    WeaviateClient: ClassVar[type] = WeaviateClient
+
+    collection_name: str = Field("mem0", description="Name of the collection")
+    embedding_model_dims: int = Field(1536, description="Dimensions of the embedding model")
+    cluster_url: Optional[str] = Field(None, description="URL for Weaviate server")
+    auth_client_secret: Optional[str] = Field(None, description="API key for Weaviate authentication")
+    additional_headers: Optional[Dict[str, str]] = Field(None, description="Additional headers for requests")
+
+    @model_validator(mode="before")
+    @classmethod
+    def check_connection_params(cls, values: Dict[str, Any]) -> Dict[str, Any]:
+        cluster_url = values.get("cluster_url")
+
+        if not cluster_url:
+            raise ValueError("'cluster_url' must be provided.")
+
+        return values
+
+    @model_validator(mode="before")
+    @classmethod
+    def validate_extra_fields(cls, values: Dict[str, Any]) -> Dict[str, Any]:
+        allowed_fields = set(cls.model_fields.keys())
+        input_fields = set(values.keys())
+        extra_fields = input_fields - allowed_fields
+
+        if extra_fields:
+            raise ValueError(
+                f"Extra fields not allowed: {', '.join(extra_fields)}. Please input only the following fields: {', '.join(allowed_fields)}"
+            )
+
+        return values
+
+    model_config = {
+        "arbitrary_types_allowed": True,
+    }

mem0/llms/anthropic.py

@@ -4,14 +4,26 @@ from typing import Dict, List, Optional
 try:
     import anthropic
 except ImportError:
-    raise ImportError("The 'anthropic' library is required. Please install it using 'pip install anthropic'.")
+    raise ImportError(
+        "The 'anthropic' library is required. Please install it using 'pip install anthropic'."
+    )
 
 from mem0.configs.llms.base import BaseLlmConfig
 from mem0.llms.base import LLMBase
 
 
 class AnthropicLLM(LLMBase):
+    """
+    A class for interacting with Anthropic's Claude models using the specified configuration.
+    """
+
     def __init__(self, config: Optional[BaseLlmConfig] = None):
+        """
+        Initializes the AnthropicLLM instance with the given configuration.
+
+        Args:
+            config (Optional[BaseLlmConfig]): Configuration settings for the language model.
+        """
         super().__init__(config)
 
         if not self.config.model:
@@ -23,23 +35,17 @@ class AnthropicLLM(LLMBase):
     def generate_response(
         self,
         messages: List[Dict[str, str]],
-        response_format=None,
-        tools: Optional[List[Dict]] = None,
-        tool_choice: str = "auto",
-    ):
+    ) -> str:
         """
-        Generate a response based on the given messages using Anthropic.
+        Generates a response using Anthropic's Claude model based on the provided messages.
 
         Args:
-            messages (list): List of message dicts containing 'role' and 'content'.
-            response_format (str or object, optional): Format of the response. Defaults to "text".
-            tools (list, optional): List of tools that the model can call. Defaults to None.
-            tool_choice (str, optional): Tool choice method. Defaults to "auto".
+            messages (List[Dict[str, str]]): A list of dictionaries, each containing a 'role' and 'content' key.
 
         Returns:
-            str: The generated response.
+            str: The generated response from the model.
         """
-        # Separate system message from other messages
+        # Extract system message separately
         system_message = ""
         filtered_messages = []
         for message in messages:
@@ -56,9 +62,6 @@ class AnthropicLLM(LLMBase):
             "max_tokens": self.config.max_tokens,
             "top_p": self.config.top_p,
         }
-        if tools:  # TODO: Remove tools if no issues found with new memory addition logic
-            params["tools"] = tools
-            params["tool_choice"] = tool_choice
 
         response = self.client.messages.create(**params)
         return response.content[0].text

mem0/llms/aws_bedrock.py

@@ -4,14 +4,26 @@ from typing import Any, Dict, List, Optional
 try:
     import boto3
 except ImportError:
-    raise ImportError("The 'boto3' library is required. Please install it using 'pip install boto3'.")
+    raise ImportError(
+        "The 'boto3' library is required. Please install it using 'pip install boto3'."
+    )
 
 from mem0.configs.llms.base import BaseLlmConfig
 from mem0.llms.base import LLMBase
 
 
 class AWSBedrockLLM(LLMBase):
+    """
+    A wrapper for AWS Bedrock's language models, integrating them with the LLMBase class.
+    """
+
     def __init__(self, config: Optional[BaseLlmConfig] = None):
+        """
+        Initializes the AWS Bedrock LLM with the provided configuration.
+
+        Args:
+            config (Optional[BaseLlmConfig]): Configuration object for the model.
+        """
         super().__init__(config)
 
         if not self.config.model:
@@ -25,49 +37,29 @@ class AWSBedrockLLM(LLMBase):
 
     def _format_messages(self, messages: List[Dict[str, str]]) -> str:
         """
-        Formats a list of messages into the required prompt structure for the model.
+        Formats a list of messages into a structured prompt for the model.
 
         Args:
-            messages (List[Dict[str, str]]): A list of dictionaries where each dictionary represents a message.
-                                            Each dictionary contains 'role' and 'content' keys.
+            messages (List[Dict[str, str]]): A list of dictionaries containing 'role' and 'content'.
 
         Returns:
             str: A formatted string combining all messages, structured with roles capitalized and separated by newlines.
         """
-        formatted_messages = []
-        for message in messages:
-            role = message["role"].capitalize()
-            content = message["content"]
-            formatted_messages.append(f"\n\n{role}: {content}")
-
+        formatted_messages = [
+            f"\n\n{msg['role'].capitalize()}: {msg['content']}" for msg in messages
+        ]
         return "".join(formatted_messages) + "\n\nAssistant:"
 
-    def _parse_response(self, response, tools) -> str:
+    def _parse_response(self, response) -> str:
         """
-        Process the response based on whether tools are used or not.
+        Extracts the generated response from the API response.
 
         Args:
-            response: The raw response from API.
-            tools: The list of tools provided in the request.
+            response: The raw response from the AWS Bedrock API.
 
         Returns:
-            str or dict: The processed response.
+            str: The generated response text.
         """
-        if tools:
-            processed_response = {"tool_calls": []}
-
-            if response["output"]["message"]["content"]:
-                for item in response["output"]["message"]["content"]:
-                    if "toolUse" in item:
-                        processed_response["tool_calls"].append(
-                            {
-                                "name": item["toolUse"]["name"],
-                                "arguments": item["toolUse"]["input"],
-                            }
-                        )
-
-            return processed_response
-
         response_body = json.loads(response["body"].read().decode())
         return response_body.get("completion", "")
 
@@ -76,22 +68,21 @@ class AWSBedrockLLM(LLMBase):
         provider: str,
         model: str,
         prompt: str,
-        model_kwargs: Optional[Dict[str, Any]] = {},
+        model_kwargs: Optional[Dict[str, Any]] = None,
     ) -> Dict[str, Any]:
         """
-        Prepares the input dictionary for the specified provider's model by mapping and renaming
-        keys in the input based on the provider's requirements.
+        Prepares the input dictionary for the specified provider's model.
 
         Args:
-            provider (str): The name of the service provider (e.g., "meta", "ai21", "mistral", "cohere", "amazon").
-            model (str): The name or identifier of the model being used.
-            prompt (str): The text prompt to be processed by the model.
-            model_kwargs (Dict[str, Any]): Additional keyword arguments specific to the model's requirements.
+            provider (str): The model provider (e.g., "meta", "ai21", "mistral", "cohere", "amazon").
+            model (str): The model identifier.
+            prompt (str): The input prompt.
+            model_kwargs (Optional[Dict[str, Any]]): Additional model parameters.
 
         Returns:
-            Dict[str, Any]: The prepared input dictionary with the correct keys and values for the specified provider.
+            Dict[str, Any]: The prepared input dictionary.
         """
-
+        model_kwargs = model_kwargs or {}
         input_body = {"prompt": prompt, **model_kwargs}
 
         provider_mappings = {
@@ -119,102 +110,35 @@ class AWSBedrockLLM(LLMBase):
                 },
             }
             input_body["textGenerationConfig"] = {
-                k: v for k, v in input_body["textGenerationConfig"].items() if v is not None
+                k: v
+                for k, v in input_body["textGenerationConfig"].items()
+                if v is not None
             }
 
         return input_body
 
-    def _convert_tool_format(self, original_tools):
+    def generate_response(self, messages: List[Dict[str, str]]) -> str:
         """
-        Converts a list of tools from their original format to a new standardized format.
+        Generates a response using AWS Bedrock based on the provided messages.
 
         Args:
-            original_tools (list): A list of dictionaries representing the original tools, each containing a 'type' key and corresponding details.
+            messages (List[Dict[str, str]]): List of message dictionaries containing 'role' and 'content'.
 
         Returns:
-            list: A list of dictionaries representing the tools in the new standardized format.
+            str: The generated response text.
         """
-        new_tools = []
+        prompt = self._format_messages(messages)
+        provider = self.config.model.split(".")[0]
+        input_body = self._prepare_input(
+            provider, self.config.model, prompt, self.model_kwargs
+        )
+        body = json.dumps(input_body)
 
-        for tool in original_tools:
-            if tool["type"] == "function":
-                function = tool["function"]
-                new_tool = {
-                    "toolSpec": {
-                        "name": function["name"],
-                        "description": function["description"],
-                        "inputSchema": {
-                            "json": {
-                                "type": "object",
-                                "properties": {},
-                                "required": function["parameters"].get("required", []),
-                            }
-                        },
-                    }
-                }
+        response = self.client.invoke_model(
+            body=body,
+            modelId=self.config.model,
+            accept="application/json",
+            contentType="application/json",
+        )
 
-                for prop, details in function["parameters"].get("properties", {}).items():
-                    new_tool["toolSpec"]["inputSchema"]["json"]["properties"][prop] = {
-                        "type": details.get("type", "string"),
-                        "description": details.get("description", ""),
-                    }
-
-                new_tools.append(new_tool)
-
-        return new_tools
-
-    def generate_response(
-        self,
-        messages: List[Dict[str, str]],
-        response_format=None,
-        tools: Optional[List[Dict]] = None,
-        tool_choice: str = "auto",
-    ):
-        """
-        Generate a response based on the given messages using AWS Bedrock.
-
-        Args:
-            messages (list): List of message dicts containing 'role' and 'content'.
-            tools (list, optional): List of tools that the model can call. Defaults to None.
-            tool_choice (str, optional): Tool choice method. Defaults to "auto".
-
-        Returns:
-            str: The generated response.
-        """
-
-        if tools:
-            # Use converse method when tools are provided
-            messages = [
-                {
-                    "role": "user",
-                    "content": [{"text": message["content"]} for message in messages],
-                }
-            ]
-            inference_config = {
-                "temperature": self.model_kwargs["temperature"],
-                "maxTokens": self.model_kwargs["max_tokens_to_sample"],
-                "topP": self.model_kwargs["top_p"],
-            }
-            tools_config = {"tools": self._convert_tool_format(tools)}
-
-            response = self.client.converse(
-                modelId=self.config.model,
-                messages=messages,
-                inferenceConfig=inference_config,
-                toolConfig=tools_config,
-            )
-        else:
-            # Use invoke_model method when no tools are provided
-            prompt = self._format_messages(messages)
-            provider = self.model.split(".")[0]
-            input_body = self._prepare_input(provider, self.config.model, prompt, **self.model_kwargs)
-            body = json.dumps(input_body)
-
-            response = self.client.invoke_model(
-                body=body,
-                modelId=self.model,
-                accept="application/json",
-                contentType="application/json",
-            )
-
-        return self._parse_response(response, tools)
+        return self._parse_response(response)

mem0/llms/azure_openai.py

@@ -9,17 +9,35 @@ from mem0.llms.base import LLMBase
 
 
 class AzureOpenAILLM(LLMBase):
+    """
+    A class for interacting with Azure OpenAI models using the specified configuration.
+    """
+
     def __init__(self, config: Optional[BaseLlmConfig] = None):
+        """
+        Initializes the AzureOpenAILLM instance with the given configuration.
+
+        Args:
+            config (Optional[BaseLlmConfig]): Configuration settings for the language model.
+        """
         super().__init__(config)
 
-        # Model name should match the custom deployment name chosen for it.
+        # Ensure model name is set; it should match the Azure OpenAI deployment name.
         if not self.config.model:
             self.config.model = "gpt-4o"
 
-        api_key = self.config.azure_kwargs.api_key or os.getenv("LLM_AZURE_OPENAI_API_KEY")
-        azure_deployment = self.config.azure_kwargs.azure_deployment or os.getenv("LLM_AZURE_DEPLOYMENT")
-        azure_endpoint = self.config.azure_kwargs.azure_endpoint or os.getenv("LLM_AZURE_ENDPOINT")
-        api_version = self.config.azure_kwargs.api_version or os.getenv("LLM_AZURE_API_VERSION")
+        api_key = self.config.azure_kwargs.api_key or os.getenv(
+            "LLM_AZURE_OPENAI_API_KEY"
+        )
+        azure_deployment = self.config.azure_kwargs.azure_deployment or os.getenv(
+            "LLM_AZURE_DEPLOYMENT"
+        )
+        azure_endpoint = self.config.azure_kwargs.azure_endpoint or os.getenv(
+            "LLM_AZURE_ENDPOINT"
+        )
+        api_version = self.config.azure_kwargs.api_version or os.getenv(
+            "LLM_AZURE_API_VERSION"
+        )
         default_headers = self.config.azure_kwargs.default_headers
 
         self.client = AzureOpenAI(
@@ -31,54 +49,20 @@ class AzureOpenAILLM(LLMBase):
             default_headers=default_headers,
         )
 
-    def _parse_response(self, response, tools):
-        """
-        Process the response based on whether tools are used or not.
-
-        Args:
-            response: The raw response from API.
-            tools: The list of tools provided in the request.
-
-        Returns:
-            str or dict: The processed response.
-        """
-        if tools:
-            processed_response = {
-                "content": response.choices[0].message.content,
-                "tool_calls": [],
-            }
-
-            if response.choices[0].message.tool_calls:
-                for tool_call in response.choices[0].message.tool_calls:
-                    processed_response["tool_calls"].append(
-                        {
-                            "name": tool_call.function.name,
-                            "arguments": json.loads(tool_call.function.arguments),
-                        }
-                    )
-
-            return processed_response
-        else:
-            return response.choices[0].message.content
-
     def generate_response(
         self,
         messages: List[Dict[str, str]],
-        response_format=None,
-        tools: Optional[List[Dict]] = None,
-        tool_choice: str = "auto",
-    ):
+        response_format: Optional[str] = None,
+    ) -> str:
         """
-        Generate a response based on the given messages using Azure OpenAI.
+        Generates a response using Azure OpenAI based on the provided messages.
 
         Args:
-            messages (list): List of message dicts containing 'role' and 'content'.
-            response_format (str or object, optional): Format of the response. Defaults to "text".
-            tools (list, optional): List of tools that the model can call. Defaults to None.
-            tool_choice (str, optional): Tool choice method. Defaults to "auto".
+            messages (List[Dict[str, str]]): A list of dictionaries, each containing a 'role' and 'content' key.
+            response_format (Optional[str]): The desired format of the response. Defaults to None.
 
         Returns:
-            str: The generated response.
+            str: The generated response from the model.
         """
         params = {
             "model": self.config.model,
@@ -87,11 +71,8 @@ class AzureOpenAILLM(LLMBase):
             "max_tokens": self.config.max_tokens,
             "top_p": self.config.top_p,
         }
+
         if response_format:
             params["response_format"] = response_format
-        if tools:  # TODO: Remove tools if no issues found with new memory addition logic
-            params["tools"] = tools
-            params["tool_choice"] = tool_choice
-
         response = self.client.chat.completions.create(**params)
-        return self._parse_response(response, tools)
+        return response.choices[0].message.content

mem0/llms/azure_openai_structured.py

@@ -9,20 +9,38 @@ from mem0.llms.base import LLMBase
 
 
 class AzureOpenAIStructuredLLM(LLMBase):
+    """
+    A class for interacting with Azure OpenAI models using the specified configuration.
+    """
+
     def __init__(self, config: Optional[BaseLlmConfig] = None):
+        """
+        Initializes the AzureOpenAIStructuredLLM instance with the given configuration.
+
+        Args:
+            config (Optional[BaseLlmConfig]): Configuration settings for the language model.
+        """
         super().__init__(config)
 
-        # Model name should match the custom deployment name chosen for it.
+        # Ensure model name is set; it should match the Azure OpenAI deployment name.
         if not self.config.model:
             self.config.model = "gpt-4o-2024-08-06"
 
-        api_key = os.getenv("LLM_AZURE_OPENAI_API_KEY") or self.config.azure_kwargs.api_key
-        azure_deployment = os.getenv("LLM_AZURE_DEPLOYMENT") or self.config.azure_kwargs.azure_deployment
-        azure_endpoint = os.getenv("LLM_AZURE_ENDPOINT") or self.config.azure_kwargs.azure_endpoint
-        api_version = os.getenv("LLM_AZURE_API_VERSION") or self.config.azure_kwargs.api_version
+        api_key = (
+            os.getenv("LLM_AZURE_OPENAI_API_KEY") or self.config.azure_kwargs.api_key
+        )
+        azure_deployment = (
+            os.getenv("LLM_AZURE_DEPLOYMENT")
+            or self.config.azure_kwargs.azure_deployment
+        )
+        azure_endpoint = (
+            os.getenv("LLM_AZURE_ENDPOINT") or self.config.azure_kwargs.azure_endpoint
+        )
+        api_version = (
+            os.getenv("LLM_AZURE_API_VERSION") or self.config.azure_kwargs.api_version
+        )
         default_headers = self.config.azure_kwargs.default_headers
 
-        # Can display a warning if API version is of model and api-version
         self.client = AzureOpenAI(
             azure_deployment=azure_deployment,
             azure_endpoint=azure_endpoint,
@@ -32,54 +50,20 @@ class AzureOpenAIStructuredLLM(LLMBase):
             default_headers=default_headers,
         )
 
-    def _parse_response(self, response, tools):
-        """
-        Process the response based on whether tools are used or not.
-
-        Args:
-            response: The raw response from API.
-            tools: The list of tools provided in the request.
-
-        Returns:
-            str or dict: The processed response.
-        """
-        if tools:
-            processed_response = {
-                "content": response.choices[0].message.content,
-                "tool_calls": [],
-            }
-
-            if response.choices[0].message.tool_calls:
-                for tool_call in response.choices[0].message.tool_calls:
-                    processed_response["tool_calls"].append(
-                        {
-                            "name": tool_call.function.name,
-                            "arguments": json.loads(tool_call.function.arguments),
-                        }
-                    )
-
-            return processed_response
-        else:
-            return response.choices[0].message.content
-
     def generate_response(
         self,
         messages: List[Dict[str, str]],
-        response_format=None,
-        tools: Optional[List[Dict]] = None,
-        tool_choice: str = "auto",
-    ):
+        response_format: Optional[str] = None,
+    ) -> str:
         """
-        Generate a response based on the given messages using Azure OpenAI.
+        Generates a response using Azure OpenAI based on the provided messages.
 
         Args:
-            messages (list): List of message dicts containing 'role' and 'content'.
-            response_format (str or object, optional): Format of the response. Defaults to "text".
-            tools (list, optional): List of tools that the model can call. Defaults to None.
-            tool_choice (str, optional): Tool choice method. Defaults to "auto".
+            messages (List[Dict[str, str]]): A list of dictionaries, each containing a 'role' and 'content' key.
+            response_format (Optional[str]): The desired format of the response. Defaults to None.
 
         Returns:
-            str: The generated response.
+            str: The generated response from the model.
         """
         params = {
             "model": self.config.model,
@@ -88,11 +72,9 @@ class AzureOpenAIStructuredLLM(LLMBase):
             "max_tokens": self.config.max_tokens,
             "top_p": self.config.top_p,
         }
+
         if response_format:
             params["response_format"] = response_format
-        if tools:
-            params["tools"] = tools
-            params["tool_choice"] = tool_choice
 
         response = self.client.chat.completions.create(**params)
-        return self._parse_response(response, tools)
+        return response.choices[0].message.content

mem0/llms/configs.py

@@ -4,8 +4,12 @@ from pydantic import BaseModel, Field, field_validator
 
 
 class LlmConfig(BaseModel):
-    provider: str = Field(description="Provider of the LLM (e.g., 'ollama', 'openai')", default="openai")
-    config: Optional[dict] = Field(description="Configuration for the specific LLM", default={})
+    provider: str = Field(
+        description="Provider of the LLM (e.g., 'ollama', 'openai')", default="openai"
+    )
+    config: Optional[dict] = Field(
+        description="Configuration for the specific LLM", default={}
+    )
 
     @field_validator("config")
     def validate_config(cls, v, values):

mem0/llms/deepseek.py

@@ -9,64 +9,42 @@ from mem0.llms.base import LLMBase
 
 
 class DeepSeekLLM(LLMBase):
+    """
+    A class for interacting with DeepSeek's language models using the specified configuration.
+    """
+
     def __init__(self, config: Optional[BaseLlmConfig] = None):
+        """
+        Initializes the DeepSeekLLM instance with the given configuration.
+
+        Args:
+            config (Optional[BaseLlmConfig]): Configuration settings for the language model.
+        """
         super().__init__(config)
 
         if not self.config.model:
             self.config.model = "deepseek-chat"
 
         api_key = self.config.api_key or os.getenv("DEEPSEEK_API_KEY")
-        base_url = self.config.deepseek_base_url or os.getenv("DEEPSEEK_API_BASE") or "https://api.deepseek.com"
+        base_url = (
+            self.config.deepseek_base_url
+            or os.getenv("DEEPSEEK_API_BASE")
+            or "https://api.deepseek.com"
+        )
         self.client = OpenAI(api_key=api_key, base_url=base_url)
 
-    def _parse_response(self, response, tools):
-        """
-        Process the response based on whether tools are used or not.
-
-        Args:
-            response: The raw response from API.
-            tools: The list of tools provided in the request.
-
-        Returns:
-            str or dict: The processed response.
-        """
-        if tools:
-            processed_response = {
-                "content": response.choices[0].message.content,
-                "tool_calls": [],
-            }
-
-            if response.choices[0].message.tool_calls:
-                for tool_call in response.choices[0].message.tool_calls:
-                    processed_response["tool_calls"].append(
-                        {
-                            "name": tool_call.function.name,
-                            "arguments": json.loads(tool_call.function.arguments),
-                        }
-                    )
-
-            return processed_response
-        else:
-            return response.choices[0].message.content
-
     def generate_response(
         self,
         messages: List[Dict[str, str]],
-        response_format=None,
-        tools: Optional[List[Dict]] = None,
-        tool_choice: str = "auto",
-    ):
+    ) -> str:
         """
-        Generate a response based on the given messages using DeepSeek.
+        Generates a response using DeepSeek based on the provided messages.
 
         Args:
-            messages (list): List of message dicts containing 'role' and 'content'.
-            response_format (str or object, optional): Format of the response. Defaults to "text".
-            tools (list, optional): List of tools that the model can call. Defaults to None.
-            tool_choice (str, optional): Tool choice method. Defaults to "auto".
+            messages (List[Dict[str, str]]): A list of dictionaries, each containing a 'role' and 'content' key.
 
         Returns:
-            str: The generated response.
+            str: The generated response from the model.
         """
         params = {
             "model": self.config.model,
@@ -75,10 +53,5 @@ class DeepSeekLLM(LLMBase):
             "max_tokens": self.config.max_tokens,
             "top_p": self.config.top_p,
         }
-
-        if tools:
-            params["tools"] = tools
-            params["tool_choice"] = tool_choice
-
         response = self.client.chat.completions.create(**params)
-        return self._parse_response(response, tools)
+        return response.choices[0].message.content

mem0/llms/gemini.py

@@ -3,8 +3,7 @@ from typing import Dict, List, Optional
 
 try:
     import google.generativeai as genai
-    from google.generativeai import GenerativeModel, protos
-    from google.generativeai.types import content_types
+    from google.generativeai import GenerativeModel
 except ImportError:
     raise ImportError(
         "The 'google-generativeai' library is required. Please install it using 'pip install google-generativeai'."
@@ -15,7 +14,17 @@ from mem0.llms.base import LLMBase
 
 
 class GeminiLLM(LLMBase):
+    """
+    A wrapper for Google's Gemini language model, integrating it with the LLMBase class.
+    """
+
     def __init__(self, config: Optional[BaseLlmConfig] = None):
+        """
+        Initializes the Gemini LLM with the provided configuration.
+
+        Args:
+            config (Optional[BaseLlmConfig]): Configuration object for the model.
+        """
         super().__init__(config)
 
         if not self.config.model:
@@ -25,51 +34,25 @@ class GeminiLLM(LLMBase):
         genai.configure(api_key=api_key)
         self.client = GenerativeModel(model_name=self.config.model)
 
-    def _parse_response(self, response, tools):
+    def _reformat_messages(
+        self, messages: List[Dict[str, str]]
+    ) -> List[Dict[str, str]]:
         """
-        Process the response based on whether tools are used or not.
+        Reformats messages to match the Gemini API's expected structure.
 
         Args:
-            response: The raw response from API.
-            tools: The list of tools provided in the request.
+            messages (List[Dict[str, str]]): A list of messages with 'role' and 'content' keys.
 
         Returns:
-            str or dict: The processed response.
-        """
-        if tools:
-            processed_response = {
-                "content": (content if (content := response.candidates[0].content.parts[0].text) else None),
-                "tool_calls": [],
-            }
-
-            for part in response.candidates[0].content.parts:
-                if fn := part.function_call:
-                    if isinstance(fn, protos.FunctionCall):
-                        fn_call = type(fn).to_dict(fn)
-                        processed_response["tool_calls"].append({"name": fn_call["name"], "arguments": fn_call["args"]})
-                        continue
-                    processed_response["tool_calls"].append({"name": fn.name, "arguments": fn.args})
-
-            return processed_response
-        else:
-            return response.candidates[0].content.parts[0].text
-
-    def _reformat_messages(self, messages: List[Dict[str, str]]):
-        """
-        Reformat messages for Gemini.
-
-        Args:
-            messages: The list of messages provided in the request.
-
-        Returns:
-            list: The list of messages in the required format.
+            List[Dict[str, str]]: Reformatted messages in the required format.
         """
         new_messages = []
 
         for message in messages:
             if message["role"] == "system":
-                content = "THIS IS A SYSTEM PROMPT. YOU MUST OBEY THIS: " + message["content"]
-
+                content = (
+                    "THIS IS A SYSTEM PROMPT. YOU MUST OBEY THIS: " + message["content"]
+                )
             else:
                 content = message["content"]
 
@@ -82,90 +65,33 @@ class GeminiLLM(LLMBase):
 
         return new_messages
 
-    def _reformat_tools(self, tools: Optional[List[Dict]]):
-        """
-        Reformat tools for Gemini.
-
-        Args:
-            tools: The list of tools provided in the request.
-
-        Returns:
-            list: The list of tools in the required format.
-        """
-
-        def remove_additional_properties(data):
-            """Recursively removes 'additionalProperties' from nested dictionaries."""
-
-            if isinstance(data, dict):
-                filtered_dict = {
-                    key: remove_additional_properties(value)
-                    for key, value in data.items()
-                    if not (key == "additionalProperties")
-                }
-                return filtered_dict
-            else:
-                return data
-
-        new_tools = []
-        if tools:
-            for tool in tools:
-                func = tool["function"].copy()
-                new_tools.append({"function_declarations": [remove_additional_properties(func)]})
-
-            # TODO: temporarily ignore it to pass tests, will come back to update according to standards later.
-            # return content_types.to_function_library(new_tools)
-
-            return new_tools
-        else:
-            return None
-
     def generate_response(
-        self,
-        messages: List[Dict[str, str]],
-        response_format=None,
-        tools: Optional[List[Dict]] = None,
-        tool_choice: str = "auto",
-    ):
+        self, messages: List[Dict[str, str]], response_format: Optional[Dict] = None
+    ) -> str:
         """
-        Generate a response based on the given messages using Gemini.
+        Generates a response from Gemini based on the given conversation history.
 
         Args:
-            messages (list): List of message dicts containing 'role' and 'content'.
-            response_format (str or object, optional): Format for the response. Defaults to "text".
-            tools (list, optional): List of tools that the model can call. Defaults to None.
-            tool_choice (str, optional): Tool choice method. Defaults to "auto".
+            messages (List[Dict[str, str]]): List of message dictionaries containing 'role' and 'content'.
+            response_format (Optional[Dict]): Specifies the response format (e.g., JSON schema).
 
         Returns:
-            str: The generated response.
+            str: The generated response as text.
         """
-
         params = {
             "temperature": self.config.temperature,
             "max_output_tokens": self.config.max_tokens,
             "top_p": self.config.top_p,
         }
 
-        if response_format is not None and response_format["type"] == "json_object":
+        if response_format and response_format.get("type") == "json_object":
             params["response_mime_type"] = "application/json"
             if "schema" in response_format:
                 params["response_schema"] = response_format["schema"]
-        if tool_choice:
-            tool_config = content_types.to_tool_config(
-                {
-                    "function_calling_config": {
-                        "mode": tool_choice,
-                        "allowed_function_names": (
-                            [tool["function"]["name"] for tool in tools] if tool_choice == "any" else None
-                        ),
-                    }
-                }
-            )
 
         response = self.client.generate_content(
             contents=self._reformat_messages(messages),
-            tools=self._reformat_tools(tools),
             generation_config=genai.GenerationConfig(**params),
-            tool_config=tool_config,
         )
 
-        return self._parse_response(response, tools)
+        return response.candidates[0].content.parts[0].text

mem0/llms/groq.py

@@ -5,14 +5,26 @@ from typing import Dict, List, Optional
 try:
     from groq import Groq
 except ImportError:
-    raise ImportError("The 'groq' library is required. Please install it using 'pip install groq'.")
+    raise ImportError(
+        "The 'groq' library is required. Please install it using 'pip install groq'."
+    )
 
 from mem0.configs.llms.base import BaseLlmConfig
 from mem0.llms.base import LLMBase
 
 
 class GroqLLM(LLMBase):
+    """
+    A class for interacting with Groq's language models using the specified configuration.
+    """
+
     def __init__(self, config: Optional[BaseLlmConfig] = None):
+        """
+        Initializes the GroqLLM instance with the given configuration.
+
+        Args:
+            config (Optional[BaseLlmConfig]): Configuration settings for the language model.
+        """
         super().__init__(config)
 
         if not self.config.model:
@@ -21,54 +33,20 @@ class GroqLLM(LLMBase):
         api_key = self.config.api_key or os.getenv("GROQ_API_KEY")
         self.client = Groq(api_key=api_key)
 
-    def _parse_response(self, response, tools):
-        """
-        Process the response based on whether tools are used or not.
-
-        Args:
-            response: The raw response from API.
-            tools: The list of tools provided in the request.
-
-        Returns:
-            str or dict: The processed response.
-        """
-        if tools:
-            processed_response = {
-                "content": response.choices[0].message.content,
-                "tool_calls": [],
-            }
-
-            if response.choices[0].message.tool_calls:
-                for tool_call in response.choices[0].message.tool_calls:
-                    processed_response["tool_calls"].append(
-                        {
-                            "name": tool_call.function.name,
-                            "arguments": json.loads(tool_call.function.arguments),
-                        }
-                    )
-
-            return processed_response
-        else:
-            return response.choices[0].message.content
-
     def generate_response(
         self,
         messages: List[Dict[str, str]],
-        response_format=None,
-        tools: Optional[List[Dict]] = None,
-        tool_choice: str = "auto",
-    ):
+        response_format: Optional[str] = None,
+    ) -> str:
         """
-        Generate a response based on the given messages using Groq.
+        Generates a response using Groq based on the provided messages.
 
         Args:
-            messages (list): List of message dicts containing 'role' and 'content'.
-            response_format (str or object, optional): Format of the response. Defaults to "text".
-            tools (list, optional): List of tools that the model can call. Defaults to None.
-            tool_choice (str, optional): Tool choice method. Defaults to "auto".
+            messages (List[Dict[str, str]]): A list of dictionaries, each containing a 'role' and 'content' key.
+            response_format (Optional[str]): The desired format of the response. Defaults to None.
 
         Returns:
-            str: The generated response.
+            str: The generated response from the model.
         """
         params = {
             "model": self.config.model,
@@ -79,9 +57,5 @@ class GroqLLM(LLMBase):
         }
         if response_format:
             params["response_format"] = response_format
-        if tools:
-            params["tools"] = tools
-            params["tool_choice"] = tool_choice
-
         response = self.client.chat.completions.create(**params)
-        return self._parse_response(response, tools)
+        return response.choices[0].message.content

mem0/llms/litellm.py

@@ -4,70 +4,50 @@ from typing import Dict, List, Optional
 try:
     import litellm
 except ImportError:
-    raise ImportError("The 'litellm' library is required. Please install it using 'pip install litellm'.")
+    raise ImportError(
+        "The 'litellm' library is required. Please install it using 'pip install litellm'."
+    )
 
 from mem0.configs.llms.base import BaseLlmConfig
 from mem0.llms.base import LLMBase
 
 
 class LiteLLM(LLMBase):
+    """
+    A class for interacting with LiteLLM's language models using the specified configuration.
+    """
+
     def __init__(self, config: Optional[BaseLlmConfig] = None):
+        """
+        Initializes the LiteLLM instance with the given configuration.
+
+        Args:
+            config (Optional[BaseLlmConfig]): Configuration settings for the language model.
+        """
         super().__init__(config)
 
         if not self.config.model:
             self.config.model = "gpt-4o-mini"
 
-    def _parse_response(self, response, tools):
-        """
-        Process the response based on whether tools are used or not.
-
-        Args:
-            response: The raw response from API.
-            tools: The list of tools provided in the request.
-
-        Returns:
-            str or dict: The processed response.
-        """
-        if tools:
-            processed_response = {
-                "content": response.choices[0].message.content,
-                "tool_calls": [],
-            }
-
-            if response.choices[0].message.tool_calls:
-                for tool_call in response.choices[0].message.tool_calls:
-                    processed_response["tool_calls"].append(
-                        {
-                            "name": tool_call.function.name,
-                            "arguments": json.loads(tool_call.function.arguments),
-                        }
-                    )
-
-            return processed_response
-        else:
-            return response.choices[0].message.content
-
     def generate_response(
         self,
         messages: List[Dict[str, str]],
-        response_format=None,
-        tools: Optional[List[Dict]] = None,
-        tool_choice: str = "auto",
-    ):
+        response_format: Optional[str] = None,
+    ) -> str:
         """
-        Generate a response based on the given messages using Litellm.
+        Generates a response using LiteLLM based on the provided messages.
 
         Args:
-            messages (list): List of message dicts containing 'role' and 'content'.
-            response_format (str or object, optional): Format of the response. Defaults to "text".
-            tools (list, optional): List of tools that the model can call. Defaults to None.
-            tool_choice (str, optional): Tool choice method. Defaults to "auto".
+            messages (List[Dict[str, str]]): A list of dictionaries, each containing a 'role' and 'content' key.
+            response_format (Optional[str]): The desired format of the response. Defaults to None.
 
         Returns:
-            str: The generated response.
+            str: The generated response from the model.
         """
         if not litellm.supports_function_calling(self.config.model):
-            raise ValueError(f"Model '{self.config.model}' in litellm does not support function calling.")
+            raise ValueError(
+                f"Model '{self.config.model}' in LiteLLM does not support function calling."
+            )
 
         params = {
             "model": self.config.model,
@@ -78,9 +58,6 @@ class LiteLLM(LLMBase):
         }
         if response_format:
             params["response_format"] = response_format
-        if tools:  # TODO: Remove tools if no issues found with new memory addition logic
-            params["tools"] = tools
-            params["tool_choice"] = tool_choice
 
         response = litellm.completion(**params)
-        return self._parse_response(response, tools)
+        return response.choices[0].message.content

mem0/llms/ollama.py

@@ -3,77 +3,56 @@ from typing import Dict, List, Optional
 try:
     from ollama import Client
 except ImportError:
-    raise ImportError("The 'ollama' library is required. Please install it using 'pip install ollama'.")
+    raise ImportError(
+        "The 'ollama' library is required. Please install it using 'pip install ollama'."
+    )
 
 from mem0.configs.llms.base import BaseLlmConfig
 from mem0.llms.base import LLMBase
 
 
 class OllamaLLM(LLMBase):
+    """
+    A class for interacting with Ollama's language models using the specified configuration.
+    """
+
     def __init__(self, config: Optional[BaseLlmConfig] = None):
+        """
+        Initializes the OllamaLLM instance with the given configuration.
+
+        Args:
+            config (Optional[BaseLlmConfig]): Configuration settings for the language model.
+        """
         super().__init__(config)
 
         if not self.config.model:
             self.config.model = "llama3.1:70b"
+
         self.client = Client(host=self.config.ollama_base_url)
         self._ensure_model_exists()
 
     def _ensure_model_exists(self):
         """
-        Ensure the specified model exists locally. If not, pull it from Ollama.
+        Ensures the specified model exists locally. If not, pulls it from Ollama.
         """
         local_models = self.client.list()["models"]
         if not any(model.get("name") == self.config.model for model in local_models):
             self.client.pull(self.config.model)
 
-    def _parse_response(self, response, tools):
-        """
-        Process the response based on whether tools are used or not.
-
-        Args:
-            response: The raw response from API.
-            tools: The list of tools provided in the request.
-
-        Returns:
-            str or dict: The processed response.
-        """
-        if tools:
-            processed_response = {
-                "content": response["message"]["content"],
-                "tool_calls": [],
-            }
-
-            if response["message"].get("tool_calls"):
-                for tool_call in response["message"]["tool_calls"]:
-                    processed_response["tool_calls"].append(
-                        {
-                            "name": tool_call["function"]["name"],
-                            "arguments": tool_call["function"]["arguments"],
-                        }
-                    )
-
-            return processed_response
-        else:
-            return response["message"]["content"]
-
     def generate_response(
         self,
         messages: List[Dict[str, str]],
-        response_format=None,
-        tools: Optional[List[Dict]] = None,
-        tool_choice: str = "auto",
-    ):
+        response_format: Optional[str] = None,
+    ) -> str:
         """
-        Generate a response based on the given messages using OpenAI.
+        Generates a response using Ollama based on the provided messages.
 
         Args:
-            messages (list): List of message dicts containing 'role' and 'content'.
-            response_format (str or object, optional): Format of the response. Defaults to "text".
-            tools (list, optional): List of tools that the model can call. Defaults to None.
-            tool_choice (str, optional): Tool choice method. Defaults to "auto".
+            messages (List[Dict[str, str]]): A list of dictionaries, each containing a 'role' and 'content' key.
+            response_format (Optional[str]): The desired format of the response. Defaults to None.
 
         Returns:
-            str: The generated response.
+            str: The generated response from the model.
         """
         params = {
             "model": self.config.model,
@@ -87,8 +66,5 @@ class OllamaLLM(LLMBase):
         if response_format:
             params["format"] = "json"
 
-        if tools:
-            params["tools"] = tools
-
         response = self.client.chat(**params)
-        return self._parse_response(response, tools)
+        return response["message"]["content"]

mem0/llms/openai.py

@@ -9,7 +9,17 @@ from mem0.llms.base import LLMBase
 
 
 class OpenAILLM(LLMBase):
+    """
+    A class to interact with OpenAI or OpenRouter APIs for generating responses using LLMs.
+    """
+
     def __init__(self, config: Optional[BaseLlmConfig] = None):
+        """
+        Initializes the OpenAILLM instance.
+
+        Args:
+            config (Optional[BaseLlmConfig]): Configuration for the LLM, including model, API key, and base URLs.
+        """
         super().__init__(config)
 
         if not self.config.model:
@@ -24,57 +34,27 @@ class OpenAILLM(LLMBase):
             )
         else:
             api_key = self.config.api_key or os.getenv("OPENAI_API_KEY")
-            base_url = self.config.openai_base_url or os.getenv("OPENAI_API_BASE") or "https://api.openai.com/v1"
+            base_url = (
+                self.config.openai_base_url
+                or os.getenv("OPENAI_API_BASE")
+                or "https://api.openai.com/v1"
+            )
             self.client = OpenAI(api_key=api_key, base_url=base_url)
 
-    def _parse_response(self, response, tools):
-        """
-        Process the response based on whether tools are used or not.
-
-        Args:
-            response: The raw response from API.
-            tools: The list of tools provided in the request.
-
-        Returns:
-            str or dict: The processed response.
-        """
-        if tools:
-            processed_response = {
-                "content": response.choices[0].message.content,
-                "tool_calls": [],
-            }
-
-            if response.choices[0].message.tool_calls:
-                for tool_call in response.choices[0].message.tool_calls:
-                    processed_response["tool_calls"].append(
-                        {
-                            "name": tool_call.function.name,
-                            "arguments": json.loads(tool_call.function.arguments),
-                        }
-                    )
-
-            return processed_response
-        else:
-            return response.choices[0].message.content
-
     def generate_response(
         self,
         messages: List[Dict[str, str]],
-        response_format=None,
-        tools: Optional[List[Dict]] = None,
-        tool_choice: str = "auto",
-    ):
+        response_format: Optional[str] = None,
+    ) -> str:
         """
-        Generate a response based on the given messages using OpenAI.
+        Generates a response based on the provided messages using OpenAI or OpenRouter.
 
         Args:
-            messages (list): List of message dicts containing 'role' and 'content'.
-            response_format (str or object, optional): Format of the response. Defaults to "text".
-            tools (list, optional): List of tools that the model can call. Defaults to None.
-            tool_choice (str, optional): Tool choice method. Defaults to "auto".
+            messages (List[Dict[str, str]]): A list of message dictionaries containing 'role' and 'content'.
+            response_format (Optional[str]): The format of the response. Defaults to None.
 
         Returns:
-            str: The generated response.
+            str: The generated response from the model.
         """
         params = {
             "model": self.config.model,
@@ -102,9 +82,6 @@ class OpenAILLM(LLMBase):
 
         if response_format:
             params["response_format"] = response_format
-        if tools:  # TODO: Remove tools if no issues found with new memory addition logic
-            params["tools"] = tools
-            params["tool_choice"] = tool_choice
 
         response = self.client.chat.completions.create(**params)
-        return self._parse_response(response, tools)
+        return response.choices[0].message.content

mem0/llms/openai_structured.py

@@ -9,66 +9,45 @@ from mem0.llms.base import LLMBase
 
 
 class OpenAIStructuredLLM(LLMBase):
+    """
+    A class for interacting with OpenAI's structured language models using the specified configuration.
+    """
+
     def __init__(self, config: Optional[BaseLlmConfig] = None):
+        """
+        Initializes the OpenAIStructuredLLM instance with the given configuration.
+
+        Args:
+            config (Optional[BaseLlmConfig]): Configuration settings for the language model.
+        """
         super().__init__(config)
 
         if not self.config.model:
             self.config.model = "gpt-4o-2024-08-06"
 
         api_key = self.config.api_key or os.getenv("OPENAI_API_KEY")
-        base_url = self.config.openai_base_url or os.getenv("OPENAI_API_BASE") or "https://api.openai.com/v1"
+        base_url = (
+            self.config.openai_base_url
+            or os.getenv("OPENAI_API_BASE")
+            or "https://api.openai.com/v1"
+        )
         self.client = OpenAI(api_key=api_key, base_url=base_url)
 
-    def _parse_response(self, response, tools):
-        """
-        Process the response based on whether tools are used or not.
-
-        Args:
-            response: The raw response from API.
-            tools (list, optional): List of tools that the model can call.
-
-        Returns:
-            str or dict: The processed response.
-        """
-
-        if tools:
-            processed_response = {
-                "content": response.choices[0].message.content,
-                "tool_calls": [],
-            }
-
-            if response.choices[0].message.tool_calls:
-                for tool_call in response.choices[0].message.tool_calls:
-                    processed_response["tool_calls"].append(
-                        {
-                            "name": tool_call.function.name,
-                            "arguments": json.loads(tool_call.function.arguments),
-                        }
-                    )
-
-            return processed_response
-
-        else:
-            return response.choices[0].message.content
-
     def generate_response(
         self,
         messages: List[Dict[str, str]],
-        response_format=None,
-        tools: Optional[List[Dict]] = None,
-        tool_choice: str = "auto",
-    ):
+        response_format: Optional[str] = None,
+    ) -> str:
         """
-        Generate a response based on the given messages using OpenAI.
+        Generates a response using OpenAI based on the provided messages.
 
         Args:
-            messages (list): List of message dicts containing 'role' and 'content'.
-            response_format (str or object, optional): Format of the response. Defaults to "text".
-            tools (list, optional): List of tools that the model can call. Defaults to None.
-            tool_choice (str, optional): Tool choice method. Defaults to "auto".
+            messages (List[Dict[str, str]]): A list of dictionaries, each containing a 'role' and 'content' key.
+            response_format (Optional[str]): The desired format of the response. Defaults to None.
+
 
         Returns:
-            str: The generated response.
+            str: The generated response from the model.
         """
         params = {
             "model": self.config.model,
@@ -78,10 +57,6 @@ class OpenAIStructuredLLM(LLMBase):
 
         if response_format:
             params["response_format"] = response_format
-        if tools:
-            params["tools"] = tools
-            params["tool_choice"] = tool_choice
 
         response = self.client.beta.chat.completions.parse(**params)
-
-        return self._parse_response(response, tools)
+        return response.choices[0].message.content

mem0/llms/together.py

@@ -5,14 +5,26 @@ from typing import Dict, List, Optional
 try:
     from together import Together
 except ImportError:
-    raise ImportError("The 'together' library is required. Please install it using 'pip install together'.")
+    raise ImportError(
+        "The 'together' library is required. Please install it using 'pip install together'."
+    )
 
 from mem0.configs.llms.base import BaseLlmConfig
 from mem0.llms.base import LLMBase
 
 
 class TogetherLLM(LLMBase):
+    """
+    A class for interacting with the TogetherAI language model using the specified configuration.
+    """
+
     def __init__(self, config: Optional[BaseLlmConfig] = None):
+        """
+        Initializes the TogetherLLM instance with the given configuration.
+
+        Args:
+            config (Optional[BaseLlmConfig]): Configuration settings for the language model.
+        """
         super().__init__(config)
 
         if not self.config.model:
@@ -21,54 +33,20 @@ class TogetherLLM(LLMBase):
         api_key = self.config.api_key or os.getenv("TOGETHER_API_KEY")
         self.client = Together(api_key=api_key)
 
-    def _parse_response(self, response, tools):
-        """
-        Process the response based on whether tools are used or not.
-
-        Args:
-            response: The raw response from API.
-            tools: The list of tools provided in the request.
-
-        Returns:
-            str or dict: The processed response.
-        """
-        if tools:
-            processed_response = {
-                "content": response.choices[0].message.content,
-                "tool_calls": [],
-            }
-
-            if response.choices[0].message.tool_calls:
-                for tool_call in response.choices[0].message.tool_calls:
-                    processed_response["tool_calls"].append(
-                        {
-                            "name": tool_call.function.name,
-                            "arguments": json.loads(tool_call.function.arguments),
-                        }
-                    )
-
-            return processed_response
-        else:
-            return response.choices[0].message.content
-
     def generate_response(
         self,
         messages: List[Dict[str, str]],
-        response_format=None,
-        tools: Optional[List[Dict]] = None,
-        tool_choice: str = "auto",
-    ):
+        response_format: Optional[str] = None,
+    ) -> str:
         """
-        Generate a response based on the given messages using TogetherAI.
+        Generates a response using TogetherAI based on the provided messages.
 
         Args:
-            messages (list): List of message dicts containing 'role' and 'content'.
-            response_format (str or object, optional): Format of the response. Defaults to "text".
-            tools (list, optional): List of tools that the model can call. Defaults to None.
-            tool_choice (str, optional): Tool choice method. Defaults to "auto".
+            messages (List[Dict[str, str]]): A list of dictionaries, each containing a 'role' and 'content' key.
+            response_format (Optional[str]): The desired format of the response. Defaults to None.
 
         Returns:
-            str: The generated response.
+            str: The generated response from the model.
         """
         params = {
             "model": self.config.model,
@@ -79,9 +57,6 @@ class TogetherLLM(LLMBase):
         }
         if response_format:
             params["response_format"] = response_format
-        if tools:  # TODO: Remove tools if no issues found with new memory addition logic
-            params["tools"] = tools
-            params["tool_choice"] = tool_choice
 
         response = self.client.chat.completions.create(**params)
-        return self._parse_response(response, tools)
+        return response.choices[0].message.content

mem0/llms/xai.py

@@ -15,7 +15,11 @@ class XAILLM(LLMBase):
             self.config.model = "grok-2-latest"
 
         api_key = self.config.api_key or os.getenv("XAI_API_KEY")
-        base_url = self.config.xai_base_url or os.getenv("XAI_API_BASE") or "https://api.x.ai/v1"
+        base_url = (
+            self.config.xai_base_url
+            or os.getenv("XAI_API_BASE")
+            or "https://api.x.ai/v1"
+        )
         self.client = OpenAI(api_key=api_key, base_url=base_url)
 
     def generate_response(self, messages: List[Dict[str, str]], response_format=None):

mem0/memory/graph_memory.py

@@ -3,23 +3,20 @@ import logging
 from mem0.memory.utils import format_entities
 
 try:
-    from langchain_community.graphs import Neo4jGraph
+    from langchain_neo4j import Neo4jGraph
 except ImportError:
-    raise ImportError("langchain_community is not installed. Please install it using pip install langchain-community")
+    raise ImportError("langchain_neo4j is not installed. Please install it using pip install langchain-neo4j")
 
 try:
     from rank_bm25 import BM25Okapi
 except ImportError:
     raise ImportError("rank_bm25 is not installed. Please install it using pip install rank-bm25")
 
-from mem0.graphs.tools import (
-    DELETE_MEMORY_STRUCT_TOOL_GRAPH,
-    DELETE_MEMORY_TOOL_GRAPH,
-    EXTRACT_ENTITIES_STRUCT_TOOL,
-    EXTRACT_ENTITIES_TOOL,
-    RELATIONS_STRUCT_TOOL,
-    RELATIONS_TOOL,
-)
+from mem0.graphs.tools import (DELETE_MEMORY_STRUCT_TOOL_GRAPH,
+                               DELETE_MEMORY_TOOL_GRAPH,
+                               EXTRACT_ENTITIES_STRUCT_TOOL,
+                               EXTRACT_ENTITIES_TOOL, RELATIONS_STRUCT_TOOL,
+                               RELATIONS_TOOL)
 from mem0.graphs.utils import EXTRACT_RELATIONS_PROMPT, get_delete_messages
 from mem0.utils.factory import EmbedderFactory, LlmFactory
 

mem0/memory/main.py

@@ -34,7 +34,8 @@ class Memory(MemoryBase):
     def __init__(self, config: MemoryConfig = MemoryConfig()):
         self.config = config
 
-        self.custom_prompt = self.config.custom_prompt
+        self.custom_fact_extraction_prompt = self.config.custom_fact_extraction_prompt
+        self.custom_update_memory_prompt = self.config.custom_update_memory_prompt
         self.embedding_model = EmbedderFactory.create(self.config.embedder.provider, self.config.embedder.config)
         self.vector_store = VectorStoreFactory.create(
             self.config.vector_store.provider, self.config.vector_store.config
@@ -57,12 +58,27 @@ class Memory(MemoryBase):
     @classmethod
     def from_config(cls, config_dict: Dict[str, Any]):
         try:
+            config = cls._process_config(config_dict)
             config = MemoryConfig(**config_dict)
         except ValidationError as e:
             logger.error(f"Configuration validation error: {e}")
             raise
         return cls(config)
 
+    @staticmethod
+    def _process_config(config_dict: Dict[str, Any]) -> Dict[str, Any]:
+        if "graph_store" in config_dict:
+            if "vector_store" not in config_dict and "embedder" in config_dict:
+                config_dict["vector_store"] = {}
+                config_dict["vector_store"]["config"] = {}
+                config_dict["vector_store"]["config"]["embedding_model_dims"] = config_dict["embedder"]["config"]["embedding_dims"]
+        try:
+            return config_dict
+        except ValidationError as e:
+            logger.error(f"Configuration validation error: {e}")
+            raise
+        
+
     def add(
         self,
         messages,
@@ -71,6 +87,7 @@ class Memory(MemoryBase):
         run_id=None,
         metadata=None,
         filters=None,
+        infer=True,
         prompt=None,
     ):
         """
@@ -83,6 +100,7 @@ class Memory(MemoryBase):
             run_id (str, optional): ID of the run creating the memory. Defaults to None.
             metadata (dict, optional): Metadata to store with the memory. Defaults to None.
             filters (dict, optional): Filters to apply to the search. Defaults to None.
+            infer (bool, optional): Whether to infer the memories. Defaults to True.
             prompt (str, optional): Prompt to use for memory deduction. Defaults to None.
 
         Returns:
@@ -121,7 +139,7 @@ class Memory(MemoryBase):
             messages = parse_vision_messages(messages)
 
         with concurrent.futures.ThreadPoolExecutor() as executor:
-            future1 = executor.submit(self._add_to_vector_store, messages, metadata, filters)
+            future1 = executor.submit(self._add_to_vector_store, messages, metadata, filters, infer)
             future2 = executor.submit(self._add_to_graph, messages, filters)
 
             concurrent.futures.wait([future1, future2])
@@ -147,11 +165,20 @@ class Memory(MemoryBase):
 
         return {"results": vector_store_result}
 
-    def _add_to_vector_store(self, messages, metadata, filters):
+    def _add_to_vector_store(self, messages, metadata, filters, infer):
+        if not infer:
+            returned_memories = []
+            for message in messages:
+                if message["role"] != "system":
+                    message_embeddings = self.embedding_model.embed(message["content"], "add")
+                    memory_id = self._create_memory(message["content"], message_embeddings, metadata)
+                    returned_memories.append({"id": memory_id, "memory": message["content"], "event": "ADD"})
+            return returned_memories
+
         parsed_messages = parse_messages(messages)
 
-        if self.custom_prompt:
-            system_prompt = self.custom_prompt
+        if self.custom_fact_extraction_prompt:
+            system_prompt = self.custom_fact_extraction_prompt
             user_prompt = f"Input:\n{parsed_messages}"
         else:
             system_prompt, user_prompt = get_fact_retrieval_messages(parsed_messages)
@@ -195,7 +222,7 @@ class Memory(MemoryBase):
             temp_uuid_mapping[str(idx)] = item["id"]
             retrieved_old_memory[idx]["id"] = str(idx)
 
-        function_calling_prompt = get_update_memory_messages(retrieved_old_memory, new_retrieved_facts)
+        function_calling_prompt = get_update_memory_messages(retrieved_old_memory, new_retrieved_facts, self.custom_update_memory_prompt)
 
         try:
             new_memories_with_actions = self.llm.generate_response(
@@ -305,15 +332,7 @@ class Memory(MemoryBase):
         ).model_dump(exclude={"score"})
 
         # Add metadata if there are additional keys
-        excluded_keys = {
-            "user_id",
-            "agent_id",
-            "run_id",
-            "hash",
-            "data",
-            "created_at",
-            "updated_at",
-        }
+        excluded_keys = {"user_id", "agent_id", "run_id", "hash", "data", "created_at", "updated_at", "id"}
         additional_metadata = {k: v for k, v in memory.payload.items() if k not in excluded_keys}
         if additional_metadata:
             memory_item["metadata"] = additional_metadata
@@ -376,6 +395,7 @@ class Memory(MemoryBase):
             "data",
             "created_at",
             "updated_at",
+            "id",
         }
         all_memories = [
             {
@@ -469,6 +489,7 @@ class Memory(MemoryBase):
             "data",
             "created_at",
             "updated_at",
+            "id",
         }
 
         original_memories = [
@@ -655,4 +676,4 @@ class Memory(MemoryBase):
         capture_event("mem0.reset", self)
 
     def chat(self, query):
-        raise NotImplementedError("Chat function not implemented yet.")
+        raise NotImplementedError("Chat function not implemented yet.")

mem0/utils/factory.py

@@ -72,6 +72,7 @@ class VectorStoreFactory:
         "vertex_ai_vector_search": "mem0.vector_stores.vertex_ai_vector_search.GoogleMatchingEngine",
         "opensearch": "mem0.vector_stores.opensearch.OpenSearchDB",
         "supabase": "mem0.vector_stores.supabase.Supabase",
+        "weaviate": "mem0.vector_stores.weaviate.Weaviate",
     }
 
     @classmethod

mem0/vector_stores/azure_ai_search.py

@@ -1,5 +1,6 @@
 import json
 import logging
+import re
 from typing import List, Optional
 
 from pydantic import BaseModel
@@ -12,6 +13,7 @@ try:
     from azure.search.documents import SearchClient
     from azure.search.documents.indexes import SearchIndexClient
     from azure.search.documents.indexes.models import (
+        BinaryQuantizationCompression,
         HnswAlgorithmConfiguration,
         ScalarQuantizationCompression,
         SearchField,
@@ -24,7 +26,7 @@ try:
     from azure.search.documents.models import VectorizedQuery
 except ImportError:
     raise ImportError(
-        "The 'azure-search-documents' library is required. Please install it using 'pip install azure-search-documents==11.5.1'."
+        "The 'azure-search-documents' library is required. Please install it using 'pip install azure-search-documents==11.5.2'."
     )
 
 logger = logging.getLogger(__name__)
@@ -37,43 +39,82 @@ class OutputData(BaseModel):
 
 
 class AzureAISearch(VectorStoreBase):
-    def __init__(self, service_name, collection_name, api_key, embedding_model_dims, use_compression):
-        """Initialize the Azure Cognitive Search vector store.
+    def __init__(
+        self,
+        service_name,
+        collection_name,
+        api_key,
+        embedding_model_dims,
+        compression_type: Optional[str] = None,  
+        use_float16: bool = False,
+    ):
+        """
+        Initialize the Azure AI Search vector store.
 
         Args:
-            service_name (str): Azure Cognitive Search service name.
+            service_name (str): Azure AI Search service name.
             collection_name (str): Index name.
-            api_key (str): API key for the Azure Cognitive Search service.
+            api_key (str): API key for the Azure AI Search service.
             embedding_model_dims (int): Dimension of the embedding vector.
-            use_compression (bool): Use scalar quantization vector compression
+            compression_type (Optional[str]): Specifies the type of quantization to use.
+                Allowed values are None (no quantization), "scalar", or "binary".
+            use_float16 (bool): Whether to store vectors in half precision (Edm.Half) or full precision (Edm.Single).
+                (Note: This flag is preserved from the initial implementation per feedback.)
         """
         self.index_name = collection_name
         self.collection_name = collection_name
         self.embedding_model_dims = embedding_model_dims
-        self.use_compression = use_compression
+        # If compression_type is None, treat it as "none".
+        self.compression_type = (compression_type or "none").lower()  
+        self.use_float16 = use_float16
+
         self.search_client = SearchClient(
             endpoint=f"https://{service_name}.search.windows.net",
             index_name=self.index_name,
             credential=AzureKeyCredential(api_key),
         )
         self.index_client = SearchIndexClient(
-            endpoint=f"https://{service_name}.search.windows.net", credential=AzureKeyCredential(api_key)
+            endpoint=f"https://{service_name}.search.windows.net",
+            credential=AzureKeyCredential(api_key),
         )
+
+        self.search_client._client._config.user_agent_policy.add_user_agent("mem0")
+        self.index_client._client._config.user_agent_policy.add_user_agent("mem0")
+
         self.create_col()  # create the collection / index
 
     def create_col(self):
-        """Create a new index in Azure Cognitive Search."""
-        vector_dimensions = self.embedding_model_dims  # Set this to the number of dimensions in your vector
-
-        if self.use_compression:
+        """Create a new index in Azure AI Search."""
+        # Determine vector type based on use_float16 setting.
+        if self.use_float16:
             vector_type = "Collection(Edm.Half)"
-            compression_name = "myCompression"
-            compression_configurations = [ScalarQuantizationCompression(compression_name=compression_name)]
         else:
             vector_type = "Collection(Edm.Single)"
-            compression_name = None
-            compression_configurations = []
 
+        # Configure compression settings based on the specified compression_type.
+        compression_configurations = []
+        compression_name = None
+        if self.compression_type == "scalar":
+            compression_name = "myCompression"
+            # For SQ, rescoring defaults to True and oversampling defaults to 4.
+            compression_configurations = [
+                ScalarQuantizationCompression(
+                    compression_name=compression_name
+                    # rescoring defaults to True and oversampling defaults to 4
+                )
+            ]
+        elif self.compression_type == "binary":
+            compression_name = "myCompression"
+            # For BQ, rescoring defaults to True and oversampling defaults to 10.
+            compression_configurations = [
+                BinaryQuantizationCompression(
+                    compression_name=compression_name
+                    # rescoring defaults to True and oversampling defaults to 10
+                )
+            ]
+        # If no compression is desired, compression_configurations remains empty.
+        
+        
         fields = [
             SimpleField(name="id", type=SearchFieldDataType.String, key=True),
             SimpleField(name="user_id", type=SearchFieldDataType.String, filterable=True),
@@ -82,8 +123,8 @@ class AzureAISearch(VectorStoreBase):
             SearchField(
                 name="vector",
                 type=vector_type,
-                searchable=True,
-                vector_search_dimensions=vector_dimensions,
+                searchable=True, 
+                vector_search_dimensions=self.embedding_model_dims,
                 vector_search_profile_name="my-vector-config",
             ),
             SimpleField(name="payload", type=SearchFieldDataType.String, searchable=True),
@@ -91,7 +132,11 @@ class AzureAISearch(VectorStoreBase):
 
         vector_search = VectorSearch(
             profiles=[
-                VectorSearchProfile(name="my-vector-config", algorithm_configuration_name="my-algorithms-config")
+                VectorSearchProfile(
+                    name="my-vector-config",
+                    algorithm_configuration_name="my-algorithms-config",
+                    compression_name=compression_name if self.compression_type != "none" else None
+                )
             ],
             algorithms=[HnswAlgorithmConfiguration(name="my-algorithms-config")],
             compressions=compression_configurations,
@@ -101,14 +146,16 @@ class AzureAISearch(VectorStoreBase):
 
     def _generate_document(self, vector, payload, id):
         document = {"id": id, "vector": vector, "payload": json.dumps(payload)}
-        # Extract additional fields if they exist
+        # Extract additional fields if they exist.
         for field in ["user_id", "run_id", "agent_id"]:
             if field in payload:
                 document[field] = payload[field]
         return document
 
+    # Note: Explicit "insert" calls may later be decoupled from memory management decisions.
     def insert(self, vectors, payloads=None, ids=None):
-        """Insert vectors into the index.
+        """
+        Insert vectors into the index.
 
         Args:
             vectors (List[List[float]]): List of vectors to insert.
@@ -116,61 +163,84 @@ class AzureAISearch(VectorStoreBase):
             ids (List[str], optional): List of IDs corresponding to vectors.
         """
         logger.info(f"Inserting {len(vectors)} vectors into index {self.index_name}")
-
         documents = [
-            self._generate_document(vector, payload, id) for id, vector, payload in zip(ids, vectors, payloads)
+            self._generate_document(vector, payload, id)
+            for id, vector, payload in zip(ids, vectors, payloads)
         ]
-        self.search_client.upload_documents(documents)
+        response = self.search_client.upload_documents(documents)
+        for doc in response:
+            if not hasattr(doc, "status_code") and doc.get("status_code") != 201:
+                raise Exception(f"Insert failed for document {doc.get('id')}: {doc}")
+        return response
+
+    def _sanitize_key(self, key: str) -> str:
+        return re.sub(r"[^\w]", "", key)
 
     def _build_filter_expression(self, filters):
         filter_conditions = []
         for key, value in filters.items():
-            # If the value is a string, add quotes
+            safe_key = self._sanitize_key(key)
             if isinstance(value, str):
-                condition = f"{key} eq '{value}'"
+                safe_value = value.replace("'", "''")
+                condition = f"{safe_key} eq '{safe_value}'"
             else:
-                condition = f"{key} eq {value}"
+                condition = f"{safe_key} eq {value}"
             filter_conditions.append(condition)
-        # Use 'and' to join multiple conditions
         filter_expression = " and ".join(filter_conditions)
         return filter_expression
 
     def search(self, query, limit=5, filters=None):
-        """Search for similar vectors.
+        """
+        Search for similar vectors.
 
         Args:
-            query (List[float]): Query vectors.
+            query (List[float]): Query vector.
             limit (int, optional): Number of results to return. Defaults to 5.
             filters (Dict, optional): Filters to apply to the search. Defaults to None.
 
         Returns:
-            list: Search results.
+            List[OutputData]: Search results.
         """
-        # Build filter expression
         filter_expression = None
         if filters:
             filter_expression = self._build_filter_expression(filters)
 
-        vector_query = VectorizedQuery(vector=query, k_nearest_neighbors=limit, fields="vector")
-        search_results = self.search_client.search(vector_queries=[vector_query], filter=filter_expression, top=limit)
+        vector_query = VectorizedQuery(
+            vector=query, k_nearest_neighbors=limit, fields="vector"
+        )
+        search_results = self.search_client.search(
+            vector_queries=[vector_query],
+            filter=filter_expression,
+            top=limit
+        )
 
         results = []
         for result in search_results:
             payload = json.loads(result["payload"])
-            results.append(OutputData(id=result["id"], score=result["@search.score"], payload=payload))
+            results.append(
+                OutputData(
+                    id=result["id"], score=result["@search.score"], payload=payload
+                )
+            )
         return results
 
     def delete(self, vector_id):
-        """Delete a vector by ID.
+        """
+        Delete a vector by ID.
 
         Args:
             vector_id (str): ID of the vector to delete.
         """
-        self.search_client.delete_documents(documents=[{"id": vector_id}])
+        response = self.search_client.delete_documents(documents=[{"id": vector_id}])
+        for doc in response:
+            if not hasattr(doc, "status_code") and doc.get("status_code") != 200:
+                raise Exception(f"Delete failed for document {vector_id}: {doc}")
         logger.info(f"Deleted document with ID '{vector_id}' from index '{self.index_name}'.")
+        return response
 
     def update(self, vector_id, vector=None, payload=None):
-        """Update a vector and its payload.
+        """
+        Update a vector and its payload.
 
         Args:
             vector_id (str): ID of the vector to update.
@@ -185,10 +255,15 @@ class AzureAISearch(VectorStoreBase):
             document["payload"] = json_payload
             for field in ["user_id", "run_id", "agent_id"]:
                 document[field] = payload.get(field)
-        self.search_client.merge_or_upload_documents(documents=[document])
+        response = self.search_client.merge_or_upload_documents(documents=[document])
+        for doc in response:
+            if not hasattr(doc, "status_code") and doc.get("status_code") != 200:
+                raise Exception(f"Update failed for document {vector_id}: {doc}")
+        return response
 
     def get(self, vector_id) -> OutputData:
-        """Retrieve a vector by ID.
+        """
+        Retrieve a vector by ID.
 
         Args:
             vector_id (str): ID of the vector to retrieve.
@@ -200,35 +275,43 @@ class AzureAISearch(VectorStoreBase):
             result = self.search_client.get_document(key=vector_id)
         except ResourceNotFoundError:
             return None
-        return OutputData(id=result["id"], score=None, payload=json.loads(result["payload"]))
+        return OutputData(
+            id=result["id"], score=None, payload=json.loads(result["payload"])
+        )
 
     def list_cols(self) -> List[str]:
-        """List all collections (indexes).
+        """
+        List all collections (indexes).
 
         Returns:
             List[str]: List of index names.
         """
-        indexes = self.index_client.list_indexes()
-        return [index.name for index in indexes]
+        try:
+            names = self.index_client.list_index_names()
+        except AttributeError:
+            names = [index.name for index in self.index_client.list_indexes()]
+        return names
 
     def delete_col(self):
         """Delete the index."""
         self.index_client.delete_index(self.index_name)
 
     def col_info(self):
-        """Get information about the index.
+        """
+        Get information about the index.
 
         Returns:
-            Dict[str, Any]: Index information.
+            dict: Index information.
         """
         index = self.index_client.get_index(self.index_name)
         return {"name": index.name, "fields": index.fields}
 
     def list(self, filters=None, limit=100):
-        """List all vectors in the index.
+        """
+        List all vectors in the index.
 
         Args:
-            filters (Dict, optional): Filters to apply to the list.
+            filters (dict, optional): Filters to apply to the list.
             limit (int, optional): Number of vectors to return. Defaults to 100.
 
         Returns:
@@ -238,12 +321,17 @@ class AzureAISearch(VectorStoreBase):
         if filters:
             filter_expression = self._build_filter_expression(filters)
 
-        search_results = self.search_client.search(search_text="*", filter=filter_expression, top=limit)
+        search_results = self.search_client.search(
+            search_text="*", filter=filter_expression, top=limit
+        )
         results = []
         for result in search_results:
             payload = json.loads(result["payload"])
-            results.append(OutputData(id=result["id"], score=result["@search.score"], payload=payload))
-
+            results.append(
+                OutputData(
+                    id=result["id"], score=result["@search.score"], payload=payload
+                )
+            )
         return [results]
 
     def __del__(self):

mem0/vector_stores/configs.py

@@ -21,6 +21,7 @@ class VectorStoreConfig(BaseModel):
         "vertex_ai_vector_search": "GoogleMatchingEngineConfig",
         "opensearch": "OpenSearchConfig",
         "supabase": "SupabaseConfig",
+        "weaviate": "WeaviateConfig",
     }
 
     @model_validator(mode="after")

mem0/vector_stores/elasticsearch.py

@@ -45,6 +45,11 @@ class ElasticsearchDB(VectorStoreBase):
         # Create index only if auto_create_index is True
         if config.auto_create_index:
             self.create_index()
+            
+        if config.custom_search_query:
+            self.custom_search_query = config.custom_search_query
+        else:
+            self.custom_search_query = None
 
     def create_index(self) -> None:
         """Create Elasticsearch index with proper mappings if it doesn't exist"""
@@ -117,25 +122,20 @@ class ElasticsearchDB(VectorStoreBase):
         return results
 
     def search(self, query: List[float], limit: int = 5, filters: Optional[Dict] = None) -> List[OutputData]:
-        """Search for similar vectors using KNN search with pre-filtering."""
-        if not filters:
-            # If no filters, just do KNN search
-            search_query = {"knn": {"field": "vector", "query_vector": query, "k": limit, "num_candidates": limit * 2}}
+        """
+        Search with two options:
+        1. Use custom search query if provided
+        2. Use KNN search on vectors with pre-filtering if no custom search query is provided
+        """
+        if self.custom_search_query:
+            search_query = self.custom_search_query(query, limit, filters)
         else:
-            # If filters exist, apply them with KNN search
-            filter_conditions = []
-            for key, value in filters.items():
-                filter_conditions.append({"term": {f"metadata.{key}": value}})
-
-            search_query = {
-                "knn": {
-                    "field": "vector",
-                    "query_vector": query,
-                    "k": limit,
-                    "num_candidates": limit * 2,
-                    "filter": {"bool": {"must": filter_conditions}},
-                }
-            }
+            search_query = {"knn": {"field": "vector", "query_vector": query, "k": limit, "num_candidates": limit * 2}}
+            if filters:
+                filter_conditions = []
+                for key, value in filters.items():
+                    filter_conditions.append({"term": {f"metadata.{key}": value}})
+                search_query["filter"] = {"bool": {"must": filter_conditions}}
 
         response = self.client.search(index=self.collection_name, body=search_query)
 

mem0/vector_stores/supabase.py

@@ -1,6 +1,6 @@
 import logging
 import uuid
-from typing import List, Optional, Dict, Any
+from typing import List, Optional
 
 from pydantic import BaseModel
 

mem0/vector_stores/weaviate.py

@@ -0,0 +1,307 @@
+import logging
+import uuid
+from typing import Dict, List, Mapping, Optional
+
+from pydantic import BaseModel
+
+try:
+    import weaviate
+except ImportError:
+    raise ImportError(
+        "The 'weaviate' library is required. Please install it using 'pip install weaviate-client weaviate'."
+    )
+
+import weaviate.classes.config as wvcc
+from weaviate.classes.init import Auth
+from weaviate.classes.query import Filter, MetadataQuery
+from weaviate.util import get_valid_uuid
+
+from mem0.vector_stores.base import VectorStoreBase
+
+logger = logging.getLogger(__name__)
+
+
+class OutputData(BaseModel):
+    id: str
+    score: float
+    payload: Dict
+
+
+class Weaviate(VectorStoreBase):
+    def __init__(
+        self,
+        collection_name: str,
+        embedding_model_dims: int,
+        cluster_url: str = None,
+        auth_client_secret: str = None,
+        additional_headers: dict = None,
+    ):
+        """
+        Initialize the Weaviate vector store.
+
+        Args:
+            collection_name (str): Name of the collection/class in Weaviate.
+            embedding_model_dims (int): Dimensions of the embedding model.
+            client (WeaviateClient, optional): Existing Weaviate client instance. Defaults to None.
+            cluster_url (str, optional): URL for Weaviate server. Defaults to None.
+            auth_config (dict, optional): Authentication configuration for Weaviate. Defaults to None.
+            additional_headers (dict, optional): Additional headers for requests. Defaults to None.
+        """
+        if "localhost" in cluster_url:
+            self.client = weaviate.connect_to_local(headers=additional_headers)
+        else:
+            self.client = weaviate.connect_to_wcs(
+                cluster_url=cluster_url,
+                auth_credentials=Auth.api_key(auth_client_secret),
+                headers=additional_headers,
+            )
+
+        self.collection_name = collection_name
+        self.create_col(embedding_model_dims)
+
+    def _parse_output(self, data: Dict) -> List[OutputData]:
+        """
+        Parse the output data.
+
+        Args:
+            data (Dict): Output data.
+
+        Returns:
+            List[OutputData]: Parsed output data.
+        """
+        keys = ["ids", "distances", "metadatas"]
+        values = []
+
+        for key in keys:
+            value = data.get(key, [])
+            if isinstance(value, list) and value and isinstance(value[0], list):
+                value = value[0]
+            values.append(value)
+
+        ids, distances, metadatas = values
+        max_length = max(len(v) for v in values if isinstance(v, list) and v is not None)
+
+        result = []
+        for i in range(max_length):
+            entry = OutputData(
+                id=ids[i] if isinstance(ids, list) and ids and i < len(ids) else None,
+                score=(distances[i] if isinstance(distances, list) and distances and i < len(distances) else None),
+                payload=(metadatas[i] if isinstance(metadatas, list) and metadatas and i < len(metadatas) else None),
+            )
+            result.append(entry)
+
+        return result
+
+    def create_col(self, vector_size, distance="cosine"):
+        """
+        Create a new collection with the specified schema.
+
+        Args:
+            vector_size (int): Size of the vectors to be stored.
+            distance (str, optional): Distance metric for vector similarity. Defaults to "cosine".
+        """
+        if self.client.collections.exists(self.collection_name):
+            logging.debug(f"Collection {self.collection_name} already exists. Skipping creation.")
+            return
+
+        properties = [
+            wvcc.Property(name="ids", data_type=wvcc.DataType.TEXT),
+            wvcc.Property(name="hash", data_type=wvcc.DataType.TEXT),
+            wvcc.Property(
+                name="metadata",
+                data_type=wvcc.DataType.TEXT,
+                description="Additional metadata",
+            ),
+            wvcc.Property(name="data", data_type=wvcc.DataType.TEXT),
+            wvcc.Property(name="created_at", data_type=wvcc.DataType.TEXT),
+            wvcc.Property(name="category", data_type=wvcc.DataType.TEXT),
+            wvcc.Property(name="updated_at", data_type=wvcc.DataType.TEXT),
+            wvcc.Property(name="user_id", data_type=wvcc.DataType.TEXT),
+            wvcc.Property(name="agent_id", data_type=wvcc.DataType.TEXT),
+            wvcc.Property(name="run_id", data_type=wvcc.DataType.TEXT),
+        ]
+
+        vectorizer_config = wvcc.Configure.Vectorizer.none()
+        vector_index_config = wvcc.Configure.VectorIndex.hnsw()
+
+        self.client.collections.create(
+            self.collection_name,
+            vectorizer_config=vectorizer_config,
+            vector_index_config=vector_index_config,
+            properties=properties,
+        )
+
+    def insert(self, vectors, payloads=None, ids=None):
+        """
+        Insert vectors into a collection.
+
+        Args:
+            vectors (list): List of vectors to insert.
+            payloads (list, optional): List of payloads corresponding to vectors. Defaults to None.
+            ids (list, optional): List of IDs corresponding to vectors. Defaults to None.
+        """
+        logger.info(f"Inserting {len(vectors)} vectors into collection {self.collection_name}")
+        with self.client.batch.fixed_size(batch_size=100) as batch:
+            for idx, vector in enumerate(vectors):
+                object_id = ids[idx] if ids and idx < len(ids) else str(uuid.uuid4())
+                object_id = get_valid_uuid(object_id)
+
+                data_object = payloads[idx] if payloads and idx < len(payloads) else {}
+
+                # Ensure 'id' is not included in properties (it's used as the Weaviate object ID)
+                if "ids" in data_object:
+                    del data_object["ids"]
+
+                batch.add_object(collection=self.collection_name, properties=data_object, uuid=object_id, vector=vector)
+
+    def search(self, query: List[float], limit: int = 5, filters: Optional[Dict] = None) -> List[OutputData]:
+        """
+        Search for similar vectors.
+        """
+        collection = self.client.collections.get(str(self.collection_name))
+        filter_conditions = []
+        if filters:
+            for key, value in filters.items():
+                if value and key in ["user_id", "agent_id", "run_id"]:
+                    filter_conditions.append(Filter.by_property(key).equal(value))
+        combined_filter = Filter.all_of(filter_conditions) if filter_conditions else None
+        response = collection.query.hybrid(
+            query="",
+            vector=query,
+            limit=limit,
+            filters=combined_filter,
+            return_properties=["hash", "created_at", "updated_at", "user_id", "agent_id", "run_id", "data", "category"],
+            return_metadata=MetadataQuery(score=True),
+        )
+        results = []
+        for obj in response.objects:
+            payload = obj.properties.copy()
+
+            for id_field in ["run_id", "agent_id", "user_id"]:
+                if id_field in payload and payload[id_field] is None:
+                    del payload[id_field]
+
+            payload["id"] = str(obj.uuid).split("'")[0]  # Include the id in the payload
+            results.append(
+                OutputData(
+                    id=str(obj.uuid),
+                    score=1
+                    if obj.metadata.distance is None
+                    else 1 - obj.metadata.distance,  # Convert distance to score
+                    payload=payload,
+                )
+            )
+        return results
+
+    def delete(self, vector_id):
+        """
+        Delete a vector by ID.
+
+        Args:
+            vector_id: ID of the vector to delete.
+        """
+        collection = self.client.collections.get(str(self.collection_name))
+        collection.data.delete_by_id(vector_id)
+
+    def update(self, vector_id, vector=None, payload=None):
+        """
+        Update a vector and its payload.
+
+        Args:
+            vector_id: ID of the vector to update.
+            vector (list, optional): Updated vector. Defaults to None.
+            payload (dict, optional): Updated payload. Defaults to None.
+        """
+        collection = self.client.collections.get(str(self.collection_name))
+
+        if payload:
+            collection.data.update(uuid=vector_id, properties=payload)
+
+        if vector:
+            existing_data = self.get(vector_id)
+            if existing_data:
+                existing_data = dict(existing_data)
+                if "id" in existing_data:
+                    del existing_data["id"]
+                existing_payload: Mapping[str, str] = existing_data
+                collection.data.update(uuid=vector_id, properties=existing_payload, vector=vector)
+
+    def get(self, vector_id):
+        """
+        Retrieve a vector by ID.
+
+        Args:
+            vector_id: ID of the vector to retrieve.
+
+        Returns:
+            dict: Retrieved vector and metadata.
+        """
+        vector_id = get_valid_uuid(vector_id)
+        collection = self.client.collections.get(str(self.collection_name))
+
+        response = collection.query.fetch_object_by_id(
+            uuid=vector_id,
+            return_properties=["hash", "created_at", "updated_at", "user_id", "agent_id", "run_id", "data", "category"],
+        )
+        # results = {}
+        # print("reponse",response)
+        # for obj in response.objects:
+        payload = response.properties.copy()
+        payload["id"] = str(response.uuid).split("'")[0]
+        results = OutputData(
+            id=str(response.uuid).split("'")[0],
+            score=1.0,
+            payload=payload,
+        )
+        return results
+
+    def list_cols(self):
+        """
+        List all collections.
+
+        Returns:
+            list: List of collection names.
+        """
+        collections = self.client.collections.list_all()
+        logger.debug(f"collections: {collections}")
+        print(f"collections: {collections}")
+        return {"collections": [{"name": col.name} for col in collections]}
+
+    def delete_col(self):
+        """Delete a collection."""
+        self.client.collections.delete(self.collection_name)
+
+    def col_info(self):
+        """
+        Get information about a collection.
+
+        Returns:
+            dict: Collection information.
+        """
+        schema = self.client.collections.get(self.collection_name)
+        if schema:
+            return schema
+        return None
+
+    def list(self, filters=None, limit=100) -> List[OutputData]:
+        """
+        List all vectors in a collection.
+        """
+        collection = self.client.collections.get(self.collection_name)
+        filter_conditions = []
+        if filters:
+            for key, value in filters.items():
+                if value and key in ["user_id", "agent_id", "run_id"]:
+                    filter_conditions.append(Filter.by_property(key).equal(value))
+        combined_filter = Filter.all_of(filter_conditions) if filter_conditions else None
+        response = collection.query.fetch_objects(
+            limit=limit,
+            filters=combined_filter,
+            return_properties=["hash", "created_at", "updated_at", "user_id", "agent_id", "run_id", "data", "category"],
+        )
+        results = []
+        for obj in response.objects:
+            payload = obj.properties.copy()
+            payload["id"] = str(obj.uuid).split("'")[0]
+            results.append(OutputData(id=str(obj.uuid).split("'")[0], score=1.0, payload=payload))
+        return [results]

poetry.lock

@@ -1,4 +1,4 @@
-# This file is automatically @generated by Poetry 2.0.1 and should not be changed by hand.
+# This file is automatically @generated by Poetry 2.1.1 and should not be changed by hand.
 
 [[package]]
 name = "aiohappyeyeballs"
@@ -7,7 +7,7 @@ description = "Happy Eyeballs for asyncio"
 optional = false
 python-versions = ">=3.9"
 groups = ["main"]
-markers = "extra == \"graph\""
+markers = "(python_version < \"3.10\" or python_version >= \"3.12\") and extra == \"graph\""
 files = [
     {file = "aiohappyeyeballs-2.4.6-py3-none-any.whl", hash = "sha256:147ec992cf873d74f5062644332c539fcd42956dc69453fe5204195e560517e1"},
     {file = "aiohappyeyeballs-2.4.6.tar.gz", hash = "sha256:9b05052f9042985d32ecbe4b59a77ae19c006a78f1344d7fdad69d28ded3d0b0"},
@@ -20,7 +20,7 @@ description = "Async http client/server framework (asyncio)"
 optional = false
 python-versions = ">=3.9"
 groups = ["main"]
-markers = "extra == \"graph\""
+markers = "(python_version < \"3.10\" or python_version >= \"3.12\") and extra == \"graph\""
 files = [
     {file = "aiohttp-3.11.13-cp310-cp310-macosx_10_9_universal2.whl", hash = "sha256:a4fe27dbbeec445e6e1291e61d61eb212ee9fed6e47998b27de71d70d3e8777d"},
     {file = "aiohttp-3.11.13-cp310-cp310-macosx_10_9_x86_64.whl", hash = "sha256:9e64ca2dbea28807f8484c13f684a2f761e69ba2640ec49dacd342763cc265ef"},
@@ -116,7 +116,7 @@ propcache = ">=0.2.0"
 yarl = ">=1.17.0,<2.0"
 
 [package.extras]
-speedups = ["Brotli", "aiodns (>=3.2.0)", "brotlicffi"]
+speedups = ["Brotli ; platform_python_implementation == \"CPython\"", "aiodns (>=3.2.0) ; sys_platform == \"linux\" or sys_platform == \"darwin\"", "brotlicffi ; platform_python_implementation != \"CPython\""]
 
 [[package]]
 name = "aiosignal"
@@ -125,7 +125,7 @@ description = "aiosignal: a list of registered asynchronous callbacks"
 optional = false
 python-versions = ">=3.9"
 groups = ["main"]
-markers = "extra == \"graph\""
+markers = "(python_version < \"3.10\" or python_version >= \"3.12\") and extra == \"graph\""
 files = [
     {file = "aiosignal-1.3.2-py2.py3-none-any.whl", hash = "sha256:45cde58e409a301715980c2b01d0c28bdde3770d8290b5eb2173759d9acb31a5"},
     {file = "aiosignal-1.3.2.tar.gz", hash = "sha256:a8c255c66fafb1e499c9351d0bf32ff2d8a0321595ebac3b93713656d2436f54"},
@@ -166,7 +166,7 @@ typing_extensions = {version = ">=4.5", markers = "python_version < \"3.13\""}
 
 [package.extras]
 doc = ["Sphinx (>=7.4,<8.0)", "packaging", "sphinx-autodoc-typehints (>=1.2.0)", "sphinx_rtd_theme"]
-test = ["anyio[trio]", "coverage[toml] (>=7)", "exceptiongroup (>=1.2.0)", "hypothesis (>=4.0)", "psutil (>=5.9)", "pytest (>=7.0)", "trustme", "truststore (>=0.9.1)", "uvloop (>=0.21)"]
+test = ["anyio[trio]", "coverage[toml] (>=7)", "exceptiongroup (>=1.2.0)", "hypothesis (>=4.0)", "psutil (>=5.9)", "pytest (>=7.0)", "trustme", "truststore (>=0.9.1) ; python_version >= \"3.10\"", "uvloop (>=0.21) ; platform_python_implementation == \"CPython\" and platform_system != \"Windows\" and python_version < \"3.14\""]
 trio = ["trio (>=0.26.1)"]
 
 [[package]]
@@ -189,19 +189,69 @@ description = "Classes Without Boilerplate"
 optional = false
 python-versions = ">=3.8"
 groups = ["main"]
-markers = "extra == \"graph\""
+markers = "(python_version < \"3.10\" or python_version >= \"3.12\") and extra == \"graph\""
 files = [
     {file = "attrs-25.1.0-py3-none-any.whl", hash = "sha256:c75a69e28a550a7e93789579c22aa26b0f5b83b75dc4e08fe092980051e1090a"},
     {file = "attrs-25.1.0.tar.gz", hash = "sha256:1c97078a80c814273a76b2a298a932eb681c87415c11dee0a6921de7f1b02c3e"},
 ]
 
 [package.extras]
-benchmark = ["cloudpickle", "hypothesis", "mypy (>=1.11.1)", "pympler", "pytest (>=4.3.0)", "pytest-codspeed", "pytest-mypy-plugins", "pytest-xdist[psutil]"]
-cov = ["cloudpickle", "coverage[toml] (>=5.3)", "hypothesis", "mypy (>=1.11.1)", "pympler", "pytest (>=4.3.0)", "pytest-mypy-plugins", "pytest-xdist[psutil]"]
-dev = ["cloudpickle", "hypothesis", "mypy (>=1.11.1)", "pre-commit-uv", "pympler", "pytest (>=4.3.0)", "pytest-mypy-plugins", "pytest-xdist[psutil]"]
+benchmark = ["cloudpickle ; platform_python_implementation == \"CPython\"", "hypothesis", "mypy (>=1.11.1) ; platform_python_implementation == \"CPython\" and python_version >= \"3.10\"", "pympler", "pytest (>=4.3.0)", "pytest-codspeed", "pytest-mypy-plugins ; platform_python_implementation == \"CPython\" and python_version >= \"3.10\"", "pytest-xdist[psutil]"]
+cov = ["cloudpickle ; platform_python_implementation == \"CPython\"", "coverage[toml] (>=5.3)", "hypothesis", "mypy (>=1.11.1) ; platform_python_implementation == \"CPython\" and python_version >= \"3.10\"", "pympler", "pytest (>=4.3.0)", "pytest-mypy-plugins ; platform_python_implementation == \"CPython\" and python_version >= \"3.10\"", "pytest-xdist[psutil]"]
+dev = ["cloudpickle ; platform_python_implementation == \"CPython\"", "hypothesis", "mypy (>=1.11.1) ; platform_python_implementation == \"CPython\" and python_version >= \"3.10\"", "pre-commit-uv", "pympler", "pytest (>=4.3.0)", "pytest-mypy-plugins ; platform_python_implementation == \"CPython\" and python_version >= \"3.10\"", "pytest-xdist[psutil]"]
 docs = ["cogapp", "furo", "myst-parser", "sphinx", "sphinx-notfound-page", "sphinxcontrib-towncrier", "towncrier (<24.7)"]
-tests = ["cloudpickle", "hypothesis", "mypy (>=1.11.1)", "pympler", "pytest (>=4.3.0)", "pytest-mypy-plugins", "pytest-xdist[psutil]"]
-tests-mypy = ["mypy (>=1.11.1)", "pytest-mypy-plugins"]
+tests = ["cloudpickle ; platform_python_implementation == \"CPython\"", "hypothesis", "mypy (>=1.11.1) ; platform_python_implementation == \"CPython\" and python_version >= \"3.10\"", "pympler", "pytest (>=4.3.0)", "pytest-mypy-plugins ; platform_python_implementation == \"CPython\" and python_version >= \"3.10\"", "pytest-xdist[psutil]"]
+tests-mypy = ["mypy (>=1.11.1) ; platform_python_implementation == \"CPython\" and python_version >= \"3.10\"", "pytest-mypy-plugins ; platform_python_implementation == \"CPython\" and python_version >= \"3.10\""]
+
+[[package]]
+name = "azure-common"
+version = "1.1.28"
+description = "Microsoft Azure Client Library for Python (Common)"
+optional = false
+python-versions = "*"
+groups = ["main"]
+files = [
+    {file = "azure-common-1.1.28.zip", hash = "sha256:4ac0cd3214e36b6a1b6a442686722a5d8cc449603aa833f3f0f40bda836704a3"},
+    {file = "azure_common-1.1.28-py2.py3-none-any.whl", hash = "sha256:5c12d3dcf4ec20599ca6b0d3e09e86e146353d443e7fcc050c9a19c1f9df20ad"},
+]
+
+[[package]]
+name = "azure-core"
+version = "1.32.0"
+description = "Microsoft Azure Core Library for Python"
+optional = false
+python-versions = ">=3.8"
+groups = ["main"]
+files = [
+    {file = "azure_core-1.32.0-py3-none-any.whl", hash = "sha256:eac191a0efb23bfa83fddf321b27b122b4ec847befa3091fa736a5c32c50d7b4"},
+    {file = "azure_core-1.32.0.tar.gz", hash = "sha256:22b3c35d6b2dae14990f6c1be2912bf23ffe50b220e708a28ab1bb92b1c730e5"},
+]
+
+[package.dependencies]
+requests = ">=2.21.0"
+six = ">=1.11.0"
+typing-extensions = ">=4.6.0"
+
+[package.extras]
+aio = ["aiohttp (>=3.0)"]
+
+[[package]]
+name = "azure-search-documents"
+version = "11.5.2"
+description = "Microsoft Azure Cognitive Search Client Library for Python"
+optional = false
+python-versions = ">=3.8"
+groups = ["main"]
+files = [
+    {file = "azure_search_documents-11.5.2-py3-none-any.whl", hash = "sha256:c949d011008a4b0bcee3db91132741b4e4d50ddb3f7e2f48944d949d4b413b11"},
+    {file = "azure_search_documents-11.5.2.tar.gz", hash = "sha256:98977dd1fa4978d3b7d8891a0856b3becb6f02cc07ff2e1ea40b9c7254ada315"},
+]
+
+[package.dependencies]
+azure-common = ">=1.1"
+azure-core = ">=1.28.0"
+isodate = ">=0.6.0"
+typing-extensions = ">=4.6.0"
 
 [[package]]
 name = "backoff"
@@ -423,23 +473,6 @@ files = [
 ]
 markers = {main = "platform_system == \"Windows\"", dev = "sys_platform == \"win32\"", test = "sys_platform == \"win32\""}
 
-[[package]]
-name = "dataclasses-json"
-version = "0.6.7"
-description = "Easily serialize dataclasses to and from JSON."
-optional = false
-python-versions = "<4.0,>=3.7"
-groups = ["main"]
-markers = "extra == \"graph\""
-files = [
-    {file = "dataclasses_json-0.6.7-py3-none-any.whl", hash = "sha256:0dbf33f26c8d5305befd61b39d2b3414e8a407bedc2834dea9b8d642666fb40a"},
-    {file = "dataclasses_json-0.6.7.tar.gz", hash = "sha256:b6b3e528266ea45b9535223bc53ca645f5208833c29229e847b3f26a1cc55fc0"},
-]
-
-[package.dependencies]
-marshmallow = ">=3.18.0,<4.0.0"
-typing-inspect = ">=0.4.0,<1"
-
 [[package]]
 name = "distro"
 version = "1.9.0"
@@ -475,7 +508,7 @@ description = "A list-like structure which implements collections.abc.MutableSeq
 optional = false
 python-versions = ">=3.8"
 groups = ["main"]
-markers = "extra == \"graph\""
+markers = "(python_version < \"3.10\" or python_version >= \"3.12\") and extra == \"graph\""
 files = [
     {file = "frozenlist-1.5.0-cp310-cp310-macosx_10_9_universal2.whl", hash = "sha256:5b6a66c18b5b9dd261ca98dffcb826a525334b2f29e7caa54e182255c5f6a65a"},
     {file = "frozenlist-1.5.0-cp310-cp310-macosx_10_9_x86_64.whl", hash = "sha256:d1b3eb7b05ea246510b43a7e53ed1653e55c2121019a97e60cad7efb881a97bb"},
@@ -571,6 +604,47 @@ files = [
     {file = "frozenlist-1.5.0.tar.gz", hash = "sha256:81d5af29e61b9c8348e876d442253723928dce6433e0e76cd925cd83f1b4b817"},
 ]
 
+[[package]]
+name = "fsspec"
+version = "2024.12.0"
+description = "File-system specification"
+optional = false
+python-versions = ">=3.8"
+groups = ["main"]
+markers = "extra == \"graph\""
+files = [
+    {file = "fsspec-2024.12.0-py3-none-any.whl", hash = "sha256:b520aed47ad9804237ff878b504267a3b0b441e97508bd6d2d8774e3db85cee2"},
+    {file = "fsspec-2024.12.0.tar.gz", hash = "sha256:670700c977ed2fb51e0d9f9253177ed20cbde4a3e5c0283cc5385b5870c8533f"},
+]
+
+[package.extras]
+abfs = ["adlfs"]
+adl = ["adlfs"]
+arrow = ["pyarrow (>=1)"]
+dask = ["dask", "distributed"]
+dev = ["pre-commit", "ruff"]
+doc = ["numpydoc", "sphinx", "sphinx-design", "sphinx-rtd-theme", "yarl"]
+dropbox = ["dropbox", "dropboxdrivefs", "requests"]
+full = ["adlfs", "aiohttp (!=4.0.0a0,!=4.0.0a1)", "dask", "distributed", "dropbox", "dropboxdrivefs", "fusepy", "gcsfs", "libarchive-c", "ocifs", "panel", "paramiko", "pyarrow (>=1)", "pygit2", "requests", "s3fs", "smbprotocol", "tqdm"]
+fuse = ["fusepy"]
+gcs = ["gcsfs"]
+git = ["pygit2"]
+github = ["requests"]
+gs = ["gcsfs"]
+gui = ["panel"]
+hdfs = ["pyarrow (>=1)"]
+http = ["aiohttp (!=4.0.0a0,!=4.0.0a1)"]
+libarchive = ["libarchive-c"]
+oci = ["ocifs"]
+s3 = ["s3fs"]
+sftp = ["paramiko"]
+smb = ["smbprotocol"]
+ssh = ["paramiko"]
+test = ["aiohttp (!=4.0.0a0,!=4.0.0a1)", "numpy", "pytest", "pytest-asyncio (!=0.22.0)", "pytest-benchmark", "pytest-cov", "pytest-mock", "pytest-recording", "pytest-rerunfailures", "requests"]
+test-downstream = ["aiobotocore (>=2.5.4,<3.0.0)", "dask-expr", "dask[dataframe,test]", "moto[server] (>4,<5)", "pytest-timeout", "xarray"]
+test-full = ["adlfs", "aiohttp (!=4.0.0a0,!=4.0.0a1)", "cloudpickle", "dask", "distributed", "dropbox", "dropboxdrivefs", "fastparquet", "fusepy", "gcsfs", "jinja2", "kerchunk", "libarchive-c", "lz4", "notebook", "numpy", "ocifs", "pandas", "panel", "paramiko", "pyarrow", "pyarrow (>=1)", "pyftpdlib", "pygit2", "pytest", "pytest-asyncio (!=0.22.0)", "pytest-benchmark", "pytest-cov", "pytest-mock", "pytest-recording", "pytest-rerunfailures", "python-snappy", "requests", "smbprotocol", "tqdm", "urllib3", "zarr", "zstandard"]
+tqdm = ["tqdm"]
+
 [[package]]
 name = "greenlet"
 version = "3.1.1"
@@ -879,25 +953,12 @@ httpcore = "==1.*"
 idna = "*"
 
 [package.extras]
-brotli = ["brotli", "brotlicffi"]
+brotli = ["brotli ; platform_python_implementation == \"CPython\"", "brotlicffi ; platform_python_implementation != \"CPython\""]
 cli = ["click (==8.*)", "pygments (==2.*)", "rich (>=10,<14)"]
 http2 = ["h2 (>=3,<5)"]
 socks = ["socksio (==1.*)"]
 zstd = ["zstandard (>=0.18.0)"]
 
-[[package]]
-name = "httpx-sse"
-version = "0.4.0"
-description = "Consume Server-Sent Event (SSE) messages with HTTPX."
-optional = false
-python-versions = ">=3.8"
-groups = ["main"]
-markers = "extra == \"graph\""
-files = [
-    {file = "httpx-sse-0.4.0.tar.gz", hash = "sha256:1e81a3a3070ce322add1d3529ed42eb5f70817f45ed6ec915ab753f961139721"},
-    {file = "httpx_sse-0.4.0-py3-none-any.whl", hash = "sha256:f329af6eae57eaa2bdfd962b42524764af68075ea87370a2de920af5341e318f"},
-]
-
 [[package]]
 name = "hyperframe"
 version = "6.1.0"
@@ -937,6 +998,18 @@ files = [
     {file = "iniconfig-2.0.0.tar.gz", hash = "sha256:2d91e135bf72d31a410b17c16da610a82cb55f6b0477d1a902134b24a455b8b3"},
 ]
 
+[[package]]
+name = "isodate"
+version = "0.7.2"
+description = "An ISO 8601 date/time/duration parser and formatter"
+optional = false
+python-versions = ">=3.7"
+groups = ["main"]
+files = [
+    {file = "isodate-0.7.2-py3-none-any.whl", hash = "sha256:28009937d8031054830160fce6d409ed342816b543597cece116d966c6d99e15"},
+    {file = "isodate-0.7.2.tar.gz", hash = "sha256:4cd1aa0f43ca76f4a6c6c0292a85f40b35ec2e43e315b59f06e6d32171a953e6"},
+]
+
 [[package]]
 name = "isort"
 version = "5.13.2"
@@ -1038,6 +1111,19 @@ files = [
     {file = "jiter-0.8.2.tar.gz", hash = "sha256:cd73d3e740666d0e639f678adb176fad25c1bcbdae88d8d7b857e1783bb4212d"},
 ]
 
+[[package]]
+name = "json-repair"
+version = "0.30.3"
+description = "A package to repair broken json strings"
+optional = false
+python-versions = ">=3.9"
+groups = ["main"]
+markers = "extra == \"graph\""
+files = [
+    {file = "json_repair-0.30.3-py3-none-any.whl", hash = "sha256:63bb588162b0958ae93d85356ecbe54c06b8c33f8a4834f93fa2719ea669804e"},
+    {file = "json_repair-0.30.3.tar.gz", hash = "sha256:0ac56e7ae9253ee9c507a7e1a3a26799c9b0bbe5e2bec1b2cc5053e90d5b05e3"},
+]
+
 [[package]]
 name = "jsonpatch"
 version = "1.33"
@@ -1074,7 +1160,7 @@ description = "Building applications with LLMs through composability"
 optional = false
 python-versions = "<4.0,>=3.9"
 groups = ["main"]
-markers = "extra == \"graph\""
+markers = "(python_version < \"3.10\" or python_version >= \"3.12\") and extra == \"graph\""
 files = [
     {file = "langchain-0.3.19-py3-none-any.whl", hash = "sha256:1e16d97db9106640b7de4c69f8f5ed22eeda56b45b9241279e83f111640eff16"},
     {file = "langchain-0.3.19.tar.gz", hash = "sha256:b96f8a445f01d15d522129ffe77cc89c8468dbd65830d153a676de8f6b899e7b"},
@@ -1114,34 +1200,44 @@ together = ["langchain-together"]
 xai = ["langchain-xai"]
 
 [[package]]
-name = "langchain-community"
-version = "0.3.18"
-description = "Community contributed LangChain integrations."
+name = "langchain"
+version = "0.3.20"
+description = "Building applications with LLMs through composability"
 optional = false
 python-versions = "<4.0,>=3.9"
 groups = ["main"]
-markers = "extra == \"graph\""
+markers = "python_version >= \"3.10\" and python_version < \"3.12\" and extra == \"graph\""
 files = [
-    {file = "langchain_community-0.3.18-py3-none-any.whl", hash = "sha256:0d4a70144a1750045c4f726f9a43379ed2484178f76e4b8295bcef3a7fdf41d5"},
-    {file = "langchain_community-0.3.18.tar.gz", hash = "sha256:fa2889a8f0b2d22b5c306fd1b070c0970e1f11b604bf55fad2f4a1d0bf68a077"},
+    {file = "langchain-0.3.20-py3-none-any.whl", hash = "sha256:273287f8e61ffdf7e811cf8799e6a71e9381325b8625fd6618900faba79cfdd0"},
+    {file = "langchain-0.3.20.tar.gz", hash = "sha256:edcc3241703e1f6557ef5a5c35cd56f9ccc25ff12e38b4829c66d94971737a93"},
 ]
 
 [package.dependencies]
-aiohttp = ">=3.8.3,<4.0.0"
-dataclasses-json = ">=0.5.7,<0.7"
-httpx-sse = ">=0.4.0,<1.0.0"
-langchain = ">=0.3.19,<1.0.0"
-langchain-core = ">=0.3.37,<1.0.0"
-langsmith = ">=0.1.125,<0.4"
-numpy = [
-    {version = ">=1.26.4,<2", markers = "python_version < \"3.12\""},
-    {version = ">=1.26.2,<3", markers = "python_version >= \"3.12\""},
-]
-pydantic-settings = ">=2.4.0,<3.0.0"
+async-timeout = {version = ">=4.0.0,<5.0.0", markers = "python_version < \"3.11\""}
+langchain-core = ">=0.3.41,<1.0.0"
+langchain-text-splitters = ">=0.3.6,<1.0.0"
+langsmith = ">=0.1.17,<0.4"
+pydantic = ">=2.7.4,<3.0.0"
 PyYAML = ">=5.3"
 requests = ">=2,<3"
 SQLAlchemy = ">=1.4,<3"
-tenacity = ">=8.1.0,<8.4.0 || >8.4.0,<10"
+
+[package.extras]
+anthropic = ["langchain-anthropic"]
+aws = ["langchain-aws"]
+cohere = ["langchain-cohere"]
+community = ["langchain-community"]
+deepseek = ["langchain-deepseek"]
+fireworks = ["langchain-fireworks"]
+google-genai = ["langchain-google-genai"]
+google-vertexai = ["langchain-google-vertexai"]
+groq = ["langchain-groq"]
+huggingface = ["langchain-huggingface"]
+mistralai = ["langchain-mistralai"]
+ollama = ["langchain-ollama"]
+openai = ["langchain-openai"]
+together = ["langchain-together"]
+xai = ["langchain-xai"]
 
 [[package]]
 name = "langchain-core"
@@ -1150,7 +1246,7 @@ description = "Building applications with LLMs through composability"
 optional = false
 python-versions = "<4.0,>=3.9"
 groups = ["main"]
-markers = "extra == \"graph\""
+markers = "(python_version < \"3.10\" or python_version >= \"3.12\") and extra == \"graph\""
 files = [
     {file = "langchain_core-0.3.40-py3-none-any.whl", hash = "sha256:9f31358741f10a13db8531e8288b8a5ae91904018c5c2e6f739d6645a98fca03"},
     {file = "langchain_core-0.3.40.tar.gz", hash = "sha256:893a238b38491967c804662c1ec7c3e6ebaf223d1125331249c3cf3862ff2746"},
@@ -1168,6 +1264,47 @@ PyYAML = ">=5.3"
 tenacity = ">=8.1.0,<8.4.0 || >8.4.0,<10.0.0"
 typing-extensions = ">=4.7"
 
+[[package]]
+name = "langchain-core"
+version = "0.3.45"
+description = "Building applications with LLMs through composability"
+optional = false
+python-versions = "<4.0,>=3.9"
+groups = ["main"]
+markers = "python_version >= \"3.10\" and python_version < \"3.12\" and extra == \"graph\""
+files = [
+    {file = "langchain_core-0.3.45-py3-none-any.whl", hash = "sha256:fe560d644c102c3f5dcfb44eb5295e26d22deab259fdd084f6b1b55a0350b77c"},
+    {file = "langchain_core-0.3.45.tar.gz", hash = "sha256:a39b8446495d1ea97311aa726478c0a13ef1d77cb7644350bad6d9d3c0141a0c"},
+]
+
+[package.dependencies]
+jsonpatch = ">=1.33,<2.0"
+langsmith = ">=0.1.125,<0.4"
+packaging = ">=23.2,<25"
+pydantic = {version = ">=2.5.2,<3.0.0", markers = "python_full_version < \"3.12.4\""}
+PyYAML = ">=5.3"
+tenacity = ">=8.1.0,<8.4.0 || >8.4.0,<10.0.0"
+typing-extensions = ">=4.7"
+
+[[package]]
+name = "langchain-neo4j"
+version = "0.4.0"
+description = "An integration package connecting Neo4j and LangChain"
+optional = false
+python-versions = "<4.0,>=3.9"
+groups = ["main"]
+markers = "extra == \"graph\""
+files = [
+    {file = "langchain_neo4j-0.4.0-py3-none-any.whl", hash = "sha256:2760b5757e7a402884cf3419830217651df97fe4f44b3fec6c96b14b6d7fd18e"},
+    {file = "langchain_neo4j-0.4.0.tar.gz", hash = "sha256:3f059a66411cec1062a2b8c44953a70d0fff9e123e9fb1d6b3f17a0bef6d6114"},
+]
+
+[package.dependencies]
+langchain = ">=0.3.7,<0.4.0"
+langchain-core = ">=0.3.8,<0.4.0"
+neo4j = ">=5.25.0,<6.0.0"
+neo4j-graphrag = ">=1.5.0,<2.0.0"
+
 [[package]]
 name = "langchain-text-splitters"
 version = "0.3.6"
@@ -1213,27 +1350,6 @@ zstandard = ">=0.23.0,<0.24.0"
 langsmith-pyo3 = ["langsmith-pyo3 (>=0.1.0rc2,<0.2.0)"]
 pytest = ["pytest (>=7.0.0)", "rich (>=13.9.4,<14.0.0)"]
 
-[[package]]
-name = "marshmallow"
-version = "3.26.1"
-description = "A lightweight library for converting complex datatypes to and from native Python datatypes."
-optional = false
-python-versions = ">=3.9"
-groups = ["main"]
-markers = "extra == \"graph\""
-files = [
-    {file = "marshmallow-3.26.1-py3-none-any.whl", hash = "sha256:3350409f20a70a7e4e11a27661187b77cdcaeb20abca41c1454fe33636bea09c"},
-    {file = "marshmallow-3.26.1.tar.gz", hash = "sha256:e6d8affb6cb61d39d26402096dc0aee12d5a26d490a121f118d2e81dc0719dc6"},
-]
-
-[package.dependencies]
-packaging = ">=17.0"
-
-[package.extras]
-dev = ["marshmallow[tests]", "pre-commit (>=3.5,<5.0)", "tox"]
-docs = ["autodocsumm (==0.2.14)", "furo (==2024.8.6)", "sphinx (==8.1.3)", "sphinx-copybutton (==0.5.2)", "sphinx-issues (==5.0.0)", "sphinxext-opengraph (==0.9.1)"]
-tests = ["pytest", "simplejson"]
-
 [[package]]
 name = "monotonic"
 version = "1.6"
@@ -1253,7 +1369,7 @@ description = "multidict implementation"
 optional = false
 python-versions = ">=3.8"
 groups = ["main"]
-markers = "extra == \"graph\""
+markers = "(python_version < \"3.10\" or python_version >= \"3.12\") and extra == \"graph\""
 files = [
     {file = "multidict-6.1.0-cp310-cp310-macosx_10_9_universal2.whl", hash = "sha256:3380252550e372e8511d49481bd836264c009adb826b23fefcc5dd3c69692f60"},
     {file = "multidict-6.1.0-cp310-cp310-macosx_10_9_x86_64.whl", hash = "sha256:99f826cbf970077383d7de805c0681799491cb939c25450b9b5b3ced03ca99f1"},
@@ -1352,19 +1468,6 @@ files = [
 [package.dependencies]
 typing-extensions = {version = ">=4.1.0", markers = "python_version < \"3.11\""}
 
-[[package]]
-name = "mypy-extensions"
-version = "1.0.0"
-description = "Type system extensions for programs checked with the mypy type checker."
-optional = false
-python-versions = ">=3.5"
-groups = ["main"]
-markers = "extra == \"graph\""
-files = [
-    {file = "mypy_extensions-1.0.0-py3-none-any.whl", hash = "sha256:4392f6c0eb8a5668a69e23d168ffa70f0be9ccfd32b5cc2d26a34ae5b844552d"},
-    {file = "mypy_extensions-1.0.0.tar.gz", hash = "sha256:75dbf8955dc00442a438fc4d0666508a9a97b6bd41aa2f0ffe9d2f2725af0782"},
-]
-
 [[package]]
 name = "neo4j"
 version = "5.28.1"
@@ -1386,6 +1489,42 @@ numpy = ["numpy (>=1.7.0,<3.0.0)"]
 pandas = ["numpy (>=1.7.0,<3.0.0)", "pandas (>=1.1.0,<3.0.0)"]
 pyarrow = ["pyarrow (>=1.0.0)"]
 
+[[package]]
+name = "neo4j-graphrag"
+version = "1.5.0"
+description = "Python package to allow easy integration to Neo4j's GraphRAG features"
+optional = false
+python-versions = "<4.0.0,>=3.9.0"
+groups = ["main"]
+markers = "extra == \"graph\""
+files = [
+    {file = "neo4j_graphrag-1.5.0-py3-none-any.whl", hash = "sha256:d778be8476aa758ff10043b373c287ada0abe90bcfb9c34e8cb4d2c4bbd20240"},
+    {file = "neo4j_graphrag-1.5.0.tar.gz", hash = "sha256:5367c08f128e83ca7028f2112f6e0b69541400adee40cdaf6e30c3563036acb3"},
+]
+
+[package.dependencies]
+fsspec = ">=2024.9.0,<2025.0.0"
+json-repair = ">=0.30.2,<0.31.0"
+neo4j = ">=5.17.0,<6.0.0"
+pydantic = ">=2.6.3,<3.0.0"
+pypdf = ">=4.3.1,<5.0.0"
+pyyaml = ">=6.0.2,<7.0.0"
+types-pyyaml = ">=6.0.12.20240917,<7.0.0.0"
+
+[package.extras]
+anthropic = ["anthropic (>=0.36.0,<0.37.0)"]
+cohere = ["cohere (>=5.9.0,<6.0.0)"]
+experimental = ["langchain-text-splitters (>=0.3.0,<0.4.0)", "llama-index (>=0.10.55,<0.11.0)", "pygraphviz (>=1.0.0,<2.0.0) ; python_version < \"3.10\"", "pygraphviz (>=1.13.0,<2.0.0) ; python_version >= \"3.10\" and python_full_version < \"4.0.0\""]
+google = ["google-cloud-aiplatform (>=1.66.0,<2.0.0)"]
+kg-creation-tools = ["pygraphviz (>=1.0.0,<2.0.0) ; python_version < \"3.10\"", "pygraphviz (>=1.13.0,<2.0.0) ; python_version >= \"3.10\" and python_full_version < \"4.0.0\""]
+mistralai = ["mistralai (>=1.0.3,<2.0.0)"]
+ollama = ["ollama (>=0.4.4,<0.5.0)"]
+openai = ["openai (>=1.51.1,<2.0.0)"]
+pinecone = ["pinecone-client (>=4.1.0,<5.0.0)"]
+qdrant = ["qdrant-client (>=1.11.3,<2.0.0)"]
+sentence-transformers = ["sentence-transformers (>=3.0.0,<4.0.0)"]
+weaviate = ["weaviate-client (>=4.6.1,<5.0.0)"]
+
 [[package]]
 name = "numpy"
 version = "1.26.4"
@@ -1393,7 +1532,7 @@ description = "Fundamental package for array computing in Python"
 optional = false
 python-versions = ">=3.9"
 groups = ["main"]
-markers = "python_version < \"3.12\""
+markers = "python_version < \"3.10\""
 files = [
     {file = "numpy-1.26.4-cp310-cp310-macosx_10_9_x86_64.whl", hash = "sha256:9ff0f4f29c51e2803569d7a51c2304de5554655a60c5d776e35b4a41413830d0"},
     {file = "numpy-1.26.4-cp310-cp310-macosx_11_0_arm64.whl", hash = "sha256:2e4ee3380d6de9c9ec04745830fd9e2eccb3e6cf790d39d7b98ffd19b0dd754a"},
@@ -1435,68 +1574,68 @@ files = [
 
 [[package]]
 name = "numpy"
-version = "2.2.3"
+version = "2.2.4"
 description = "Fundamental package for array computing in Python"
 optional = false
 python-versions = ">=3.10"
 groups = ["main"]
-markers = "python_version >= \"3.12\""
+markers = "python_version >= \"3.10\""
 files = [
-    {file = "numpy-2.2.3-cp310-cp310-macosx_10_9_x86_64.whl", hash = "sha256:cbc6472e01952d3d1b2772b720428f8b90e2deea8344e854df22b0618e9cce71"},
-    {file = "numpy-2.2.3-cp310-cp310-macosx_11_0_arm64.whl", hash = "sha256:cdfe0c22692a30cd830c0755746473ae66c4a8f2e7bd508b35fb3b6a0813d787"},
-    {file = "numpy-2.2.3-cp310-cp310-macosx_14_0_arm64.whl", hash = "sha256:e37242f5324ffd9f7ba5acf96d774f9276aa62a966c0bad8dae692deebec7716"},
-    {file = "numpy-2.2.3-cp310-cp310-macosx_14_0_x86_64.whl", hash = "sha256:95172a21038c9b423e68be78fd0be6e1b97674cde269b76fe269a5dfa6fadf0b"},
-    {file = "numpy-2.2.3-cp310-cp310-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:d5b47c440210c5d1d67e1cf434124e0b5c395eee1f5806fdd89b553ed1acd0a3"},
-    {file = "numpy-2.2.3-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:0391ea3622f5c51a2e29708877d56e3d276827ac5447d7f45e9bc4ade8923c52"},
-    {file = "numpy-2.2.3-cp310-cp310-musllinux_1_2_aarch64.whl", hash = "sha256:f6b3dfc7661f8842babd8ea07e9897fe3d9b69a1d7e5fbb743e4160f9387833b"},
-    {file = "numpy-2.2.3-cp310-cp310-musllinux_1_2_x86_64.whl", hash = "sha256:1ad78ce7f18ce4e7df1b2ea4019b5817a2f6a8a16e34ff2775f646adce0a5027"},
-    {file = "numpy-2.2.3-cp310-cp310-win32.whl", hash = "sha256:5ebeb7ef54a7be11044c33a17b2624abe4307a75893c001a4800857956b41094"},
-    {file = "numpy-2.2.3-cp310-cp310-win_amd64.whl", hash = "sha256:596140185c7fa113563c67c2e894eabe0daea18cf8e33851738c19f70ce86aeb"},
-    {file = "numpy-2.2.3-cp311-cp311-macosx_10_9_x86_64.whl", hash = "sha256:16372619ee728ed67a2a606a614f56d3eabc5b86f8b615c79d01957062826ca8"},
-    {file = "numpy-2.2.3-cp311-cp311-macosx_11_0_arm64.whl", hash = "sha256:5521a06a3148686d9269c53b09f7d399a5725c47bbb5b35747e1cb76326b714b"},
-    {file = "numpy-2.2.3-cp311-cp311-macosx_14_0_arm64.whl", hash = "sha256:7c8dde0ca2f77828815fd1aedfdf52e59071a5bae30dac3b4da2a335c672149a"},
-    {file = "numpy-2.2.3-cp311-cp311-macosx_14_0_x86_64.whl", hash = "sha256:77974aba6c1bc26e3c205c2214f0d5b4305bdc719268b93e768ddb17e3fdd636"},
-    {file = "numpy-2.2.3-cp311-cp311-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:d42f9c36d06440e34226e8bd65ff065ca0963aeecada587b937011efa02cdc9d"},
-    {file = "numpy-2.2.3-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:f2712c5179f40af9ddc8f6727f2bd910ea0eb50206daea75f58ddd9fa3f715bb"},
-    {file = "numpy-2.2.3-cp311-cp311-musllinux_1_2_aarch64.whl", hash = "sha256:c8b0451d2ec95010d1db8ca733afc41f659f425b7f608af569711097fd6014e2"},
-    {file = "numpy-2.2.3-cp311-cp311-musllinux_1_2_x86_64.whl", hash = "sha256:d9b4a8148c57ecac25a16b0e11798cbe88edf5237b0df99973687dd866f05e1b"},
-    {file = "numpy-2.2.3-cp311-cp311-win32.whl", hash = "sha256:1f45315b2dc58d8a3e7754fe4e38b6fce132dab284a92851e41b2b344f6441c5"},
-    {file = "numpy-2.2.3-cp311-cp311-win_amd64.whl", hash = "sha256:9f48ba6f6c13e5e49f3d3efb1b51c8193215c42ac82610a04624906a9270be6f"},
-    {file = "numpy-2.2.3-cp312-cp312-macosx_10_13_x86_64.whl", hash = "sha256:12c045f43b1d2915eca6b880a7f4a256f59d62df4f044788c8ba67709412128d"},
-    {file = "numpy-2.2.3-cp312-cp312-macosx_11_0_arm64.whl", hash = "sha256:87eed225fd415bbae787f93a457af7f5990b92a334e346f72070bf569b9c9c95"},
-    {file = "numpy-2.2.3-cp312-cp312-macosx_14_0_arm64.whl", hash = "sha256:712a64103d97c404e87d4d7c47fb0c7ff9acccc625ca2002848e0d53288b90ea"},
-    {file = "numpy-2.2.3-cp312-cp312-macosx_14_0_x86_64.whl", hash = "sha256:a5ae282abe60a2db0fd407072aff4599c279bcd6e9a2475500fc35b00a57c532"},
-    {file = "numpy-2.2.3-cp312-cp312-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:5266de33d4c3420973cf9ae3b98b54a2a6d53a559310e3236c4b2b06b9c07d4e"},
-    {file = "numpy-2.2.3-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:3b787adbf04b0db1967798dba8da1af07e387908ed1553a0d6e74c084d1ceafe"},
-    {file = "numpy-2.2.3-cp312-cp312-musllinux_1_2_aarch64.whl", hash = "sha256:34c1b7e83f94f3b564b35f480f5652a47007dd91f7c839f404d03279cc8dd021"},
-    {file = "numpy-2.2.3-cp312-cp312-musllinux_1_2_x86_64.whl", hash = "sha256:4d8335b5f1b6e2bce120d55fb17064b0262ff29b459e8493d1785c18ae2553b8"},
-    {file = "numpy-2.2.3-cp312-cp312-win32.whl", hash = "sha256:4d9828d25fb246bedd31e04c9e75714a4087211ac348cb39c8c5f99dbb6683fe"},
-    {file = "numpy-2.2.3-cp312-cp312-win_amd64.whl", hash = "sha256:83807d445817326b4bcdaaaf8e8e9f1753da04341eceec705c001ff342002e5d"},
-    {file = "numpy-2.2.3-cp313-cp313-macosx_10_13_x86_64.whl", hash = "sha256:7bfdb06b395385ea9b91bf55c1adf1b297c9fdb531552845ff1d3ea6e40d5aba"},
-    {file = "numpy-2.2.3-cp313-cp313-macosx_11_0_arm64.whl", hash = "sha256:23c9f4edbf4c065fddb10a4f6e8b6a244342d95966a48820c614891e5059bb50"},
-    {file = "numpy-2.2.3-cp313-cp313-macosx_14_0_arm64.whl", hash = "sha256:a0c03b6be48aaf92525cccf393265e02773be8fd9551a2f9adbe7db1fa2b60f1"},
-    {file = "numpy-2.2.3-cp313-cp313-macosx_14_0_x86_64.whl", hash = "sha256:2376e317111daa0a6739e50f7ee2a6353f768489102308b0d98fcf4a04f7f3b5"},
-    {file = "numpy-2.2.3-cp313-cp313-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:8fb62fe3d206d72fe1cfe31c4a1106ad2b136fcc1606093aeab314f02930fdf2"},
-    {file = "numpy-2.2.3-cp313-cp313-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:52659ad2534427dffcc36aac76bebdd02b67e3b7a619ac67543bc9bfe6b7cdb1"},
-    {file = "numpy-2.2.3-cp313-cp313-musllinux_1_2_aarch64.whl", hash = "sha256:1b416af7d0ed3271cad0f0a0d0bee0911ed7eba23e66f8424d9f3dfcdcae1304"},
-    {file = "numpy-2.2.3-cp313-cp313-musllinux_1_2_x86_64.whl", hash = "sha256:1402da8e0f435991983d0a9708b779f95a8c98c6b18a171b9f1be09005e64d9d"},
-    {file = "numpy-2.2.3-cp313-cp313-win32.whl", hash = "sha256:136553f123ee2951bfcfbc264acd34a2fc2f29d7cdf610ce7daf672b6fbaa693"},
-    {file = "numpy-2.2.3-cp313-cp313-win_amd64.whl", hash = "sha256:5b732c8beef1d7bc2d9e476dbba20aaff6167bf205ad9aa8d30913859e82884b"},
-    {file = "numpy-2.2.3-cp313-cp313t-macosx_10_13_x86_64.whl", hash = "sha256:435e7a933b9fda8126130b046975a968cc2d833b505475e588339e09f7672890"},
-    {file = "numpy-2.2.3-cp313-cp313t-macosx_11_0_arm64.whl", hash = "sha256:7678556eeb0152cbd1522b684dcd215250885993dd00adb93679ec3c0e6e091c"},
-    {file = "numpy-2.2.3-cp313-cp313t-macosx_14_0_arm64.whl", hash = "sha256:2e8da03bd561504d9b20e7a12340870dfc206c64ea59b4cfee9fceb95070ee94"},
-    {file = "numpy-2.2.3-cp313-cp313t-macosx_14_0_x86_64.whl", hash = "sha256:c9aa4496fd0e17e3843399f533d62857cef5900facf93e735ef65aa4bbc90ef0"},
-    {file = "numpy-2.2.3-cp313-cp313t-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:f4ca91d61a4bf61b0f2228f24bbfa6a9facd5f8af03759fe2a655c50ae2c6610"},
-    {file = "numpy-2.2.3-cp313-cp313t-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:deaa09cd492e24fd9b15296844c0ad1b3c976da7907e1c1ed3a0ad21dded6f76"},
-    {file = "numpy-2.2.3-cp313-cp313t-musllinux_1_2_aarch64.whl", hash = "sha256:246535e2f7496b7ac85deffe932896a3577be7af8fb7eebe7146444680297e9a"},
-    {file = "numpy-2.2.3-cp313-cp313t-musllinux_1_2_x86_64.whl", hash = "sha256:daf43a3d1ea699402c5a850e5313680ac355b4adc9770cd5cfc2940e7861f1bf"},
-    {file = "numpy-2.2.3-cp313-cp313t-win32.whl", hash = "sha256:cf802eef1f0134afb81fef94020351be4fe1d6681aadf9c5e862af6602af64ef"},
-    {file = "numpy-2.2.3-cp313-cp313t-win_amd64.whl", hash = "sha256:aee2512827ceb6d7f517c8b85aa5d3923afe8fc7a57d028cffcd522f1c6fd082"},
-    {file = "numpy-2.2.3-pp310-pypy310_pp73-macosx_10_15_x86_64.whl", hash = "sha256:3c2ec8a0f51d60f1e9c0c5ab116b7fc104b165ada3f6c58abf881cb2eb16044d"},
-    {file = "numpy-2.2.3-pp310-pypy310_pp73-macosx_14_0_x86_64.whl", hash = "sha256:ed2cf9ed4e8ebc3b754d398cba12f24359f018b416c380f577bbae112ca52fc9"},
-    {file = "numpy-2.2.3-pp310-pypy310_pp73-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:39261798d208c3095ae4f7bc8eaeb3481ea8c6e03dc48028057d3cbdbdb8937e"},
-    {file = "numpy-2.2.3-pp310-pypy310_pp73-win_amd64.whl", hash = "sha256:783145835458e60fa97afac25d511d00a1eca94d4a8f3ace9fe2043003c678e4"},
-    {file = "numpy-2.2.3.tar.gz", hash = "sha256:dbdc15f0c81611925f382dfa97b3bd0bc2c1ce19d4fe50482cb0ddc12ba30020"},
+    {file = "numpy-2.2.4-cp310-cp310-macosx_10_9_x86_64.whl", hash = "sha256:8146f3550d627252269ac42ae660281d673eb6f8b32f113538e0cc2a9aed42b9"},
+    {file = "numpy-2.2.4-cp310-cp310-macosx_11_0_arm64.whl", hash = "sha256:e642d86b8f956098b564a45e6f6ce68a22c2c97a04f5acd3f221f57b8cb850ae"},
+    {file = "numpy-2.2.4-cp310-cp310-macosx_14_0_arm64.whl", hash = "sha256:a84eda42bd12edc36eb5b53bbcc9b406820d3353f1994b6cfe453a33ff101775"},
+    {file = "numpy-2.2.4-cp310-cp310-macosx_14_0_x86_64.whl", hash = "sha256:4ba5054787e89c59c593a4169830ab362ac2bee8a969249dc56e5d7d20ff8df9"},
+    {file = "numpy-2.2.4-cp310-cp310-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:7716e4a9b7af82c06a2543c53ca476fa0b57e4d760481273e09da04b74ee6ee2"},
+    {file = "numpy-2.2.4-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:adf8c1d66f432ce577d0197dceaac2ac00c0759f573f28516246351c58a85020"},
+    {file = "numpy-2.2.4-cp310-cp310-musllinux_1_2_aarch64.whl", hash = "sha256:218f061d2faa73621fa23d6359442b0fc658d5b9a70801373625d958259eaca3"},
+    {file = "numpy-2.2.4-cp310-cp310-musllinux_1_2_x86_64.whl", hash = "sha256:df2f57871a96bbc1b69733cd4c51dc33bea66146b8c63cacbfed73eec0883017"},
+    {file = "numpy-2.2.4-cp310-cp310-win32.whl", hash = "sha256:a0258ad1f44f138b791327961caedffbf9612bfa504ab9597157806faa95194a"},
+    {file = "numpy-2.2.4-cp310-cp310-win_amd64.whl", hash = "sha256:0d54974f9cf14acf49c60f0f7f4084b6579d24d439453d5fc5805d46a165b542"},
+    {file = "numpy-2.2.4-cp311-cp311-macosx_10_9_x86_64.whl", hash = "sha256:e9e0a277bb2eb5d8a7407e14688b85fd8ad628ee4e0c7930415687b6564207a4"},
+    {file = "numpy-2.2.4-cp311-cp311-macosx_11_0_arm64.whl", hash = "sha256:9eeea959168ea555e556b8188da5fa7831e21d91ce031e95ce23747b7609f8a4"},
+    {file = "numpy-2.2.4-cp311-cp311-macosx_14_0_arm64.whl", hash = "sha256:bd3ad3b0a40e713fc68f99ecfd07124195333f1e689387c180813f0e94309d6f"},
+    {file = "numpy-2.2.4-cp311-cp311-macosx_14_0_x86_64.whl", hash = "sha256:cf28633d64294969c019c6df4ff37f5698e8326db68cc2b66576a51fad634880"},
+    {file = "numpy-2.2.4-cp311-cp311-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:2fa8fa7697ad1646b5c93de1719965844e004fcad23c91228aca1cf0800044a1"},
+    {file = "numpy-2.2.4-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:f4162988a360a29af158aeb4a2f4f09ffed6a969c9776f8f3bdee9b06a8ab7e5"},
+    {file = "numpy-2.2.4-cp311-cp311-musllinux_1_2_aarch64.whl", hash = "sha256:892c10d6a73e0f14935c31229e03325a7b3093fafd6ce0af704be7f894d95687"},
+    {file = "numpy-2.2.4-cp311-cp311-musllinux_1_2_x86_64.whl", hash = "sha256:db1f1c22173ac1c58db249ae48aa7ead29f534b9a948bc56828337aa84a32ed6"},
+    {file = "numpy-2.2.4-cp311-cp311-win32.whl", hash = "sha256:ea2bb7e2ae9e37d96835b3576a4fa4b3a97592fbea8ef7c3587078b0068b8f09"},
+    {file = "numpy-2.2.4-cp311-cp311-win_amd64.whl", hash = "sha256:f7de08cbe5551911886d1ab60de58448c6df0f67d9feb7d1fb21e9875ef95e91"},
+    {file = "numpy-2.2.4-cp312-cp312-macosx_10_13_x86_64.whl", hash = "sha256:a7b9084668aa0f64e64bd00d27ba5146ef1c3a8835f3bd912e7a9e01326804c4"},
+    {file = "numpy-2.2.4-cp312-cp312-macosx_11_0_arm64.whl", hash = "sha256:dbe512c511956b893d2dacd007d955a3f03d555ae05cfa3ff1c1ff6df8851854"},
+    {file = "numpy-2.2.4-cp312-cp312-macosx_14_0_arm64.whl", hash = "sha256:bb649f8b207ab07caebba230d851b579a3c8711a851d29efe15008e31bb4de24"},
+    {file = "numpy-2.2.4-cp312-cp312-macosx_14_0_x86_64.whl", hash = "sha256:f34dc300df798742b3d06515aa2a0aee20941c13579d7a2f2e10af01ae4901ee"},
+    {file = "numpy-2.2.4-cp312-cp312-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:c3f7ac96b16955634e223b579a3e5798df59007ca43e8d451a0e6a50f6bfdfba"},
+    {file = "numpy-2.2.4-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:4f92084defa704deadd4e0a5ab1dc52d8ac9e8a8ef617f3fbb853e79b0ea3592"},
+    {file = "numpy-2.2.4-cp312-cp312-musllinux_1_2_aarch64.whl", hash = "sha256:7a4e84a6283b36632e2a5b56e121961f6542ab886bc9e12f8f9818b3c266bfbb"},
+    {file = "numpy-2.2.4-cp312-cp312-musllinux_1_2_x86_64.whl", hash = "sha256:11c43995255eb4127115956495f43e9343736edb7fcdb0d973defd9de14cd84f"},
+    {file = "numpy-2.2.4-cp312-cp312-win32.whl", hash = "sha256:65ef3468b53269eb5fdb3a5c09508c032b793da03251d5f8722b1194f1790c00"},
+    {file = "numpy-2.2.4-cp312-cp312-win_amd64.whl", hash = "sha256:2aad3c17ed2ff455b8eaafe06bcdae0062a1db77cb99f4b9cbb5f4ecb13c5146"},
+    {file = "numpy-2.2.4-cp313-cp313-macosx_10_13_x86_64.whl", hash = "sha256:1cf4e5c6a278d620dee9ddeb487dc6a860f9b199eadeecc567f777daace1e9e7"},
+    {file = "numpy-2.2.4-cp313-cp313-macosx_11_0_arm64.whl", hash = "sha256:1974afec0b479e50438fc3648974268f972e2d908ddb6d7fb634598cdb8260a0"},
+    {file = "numpy-2.2.4-cp313-cp313-macosx_14_0_arm64.whl", hash = "sha256:79bd5f0a02aa16808fcbc79a9a376a147cc1045f7dfe44c6e7d53fa8b8a79392"},
+    {file = "numpy-2.2.4-cp313-cp313-macosx_14_0_x86_64.whl", hash = "sha256:3387dd7232804b341165cedcb90694565a6015433ee076c6754775e85d86f1fc"},
+    {file = "numpy-2.2.4-cp313-cp313-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:6f527d8fdb0286fd2fd97a2a96c6be17ba4232da346931d967a0630050dfd298"},
+    {file = "numpy-2.2.4-cp313-cp313-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:bce43e386c16898b91e162e5baaad90c4b06f9dcbe36282490032cec98dc8ae7"},
+    {file = "numpy-2.2.4-cp313-cp313-musllinux_1_2_aarch64.whl", hash = "sha256:31504f970f563d99f71a3512d0c01a645b692b12a63630d6aafa0939e52361e6"},
+    {file = "numpy-2.2.4-cp313-cp313-musllinux_1_2_x86_64.whl", hash = "sha256:81413336ef121a6ba746892fad881a83351ee3e1e4011f52e97fba79233611fd"},
+    {file = "numpy-2.2.4-cp313-cp313-win32.whl", hash = "sha256:f486038e44caa08dbd97275a9a35a283a8f1d2f0ee60ac260a1790e76660833c"},
+    {file = "numpy-2.2.4-cp313-cp313-win_amd64.whl", hash = "sha256:207a2b8441cc8b6a2a78c9ddc64d00d20c303d79fba08c577752f080c4007ee3"},
+    {file = "numpy-2.2.4-cp313-cp313t-macosx_10_13_x86_64.whl", hash = "sha256:8120575cb4882318c791f839a4fd66161a6fa46f3f0a5e613071aae35b5dd8f8"},
+    {file = "numpy-2.2.4-cp313-cp313t-macosx_11_0_arm64.whl", hash = "sha256:a761ba0fa886a7bb33c6c8f6f20213735cb19642c580a931c625ee377ee8bd39"},
+    {file = "numpy-2.2.4-cp313-cp313t-macosx_14_0_arm64.whl", hash = "sha256:ac0280f1ba4a4bfff363a99a6aceed4f8e123f8a9b234c89140f5e894e452ecd"},
+    {file = "numpy-2.2.4-cp313-cp313t-macosx_14_0_x86_64.whl", hash = "sha256:879cf3a9a2b53a4672a168c21375166171bc3932b7e21f622201811c43cdd3b0"},
+    {file = "numpy-2.2.4-cp313-cp313t-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:f05d4198c1bacc9124018109c5fba2f3201dbe7ab6e92ff100494f236209c960"},
+    {file = "numpy-2.2.4-cp313-cp313t-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:e2f085ce2e813a50dfd0e01fbfc0c12bbe5d2063d99f8b29da30e544fb6483b8"},
+    {file = "numpy-2.2.4-cp313-cp313t-musllinux_1_2_aarch64.whl", hash = "sha256:92bda934a791c01d6d9d8e038363c50918ef7c40601552a58ac84c9613a665bc"},
+    {file = "numpy-2.2.4-cp313-cp313t-musllinux_1_2_x86_64.whl", hash = "sha256:ee4d528022f4c5ff67332469e10efe06a267e32f4067dc76bb7e2cddf3cd25ff"},
+    {file = "numpy-2.2.4-cp313-cp313t-win32.whl", hash = "sha256:05c076d531e9998e7e694c36e8b349969c56eadd2cdcd07242958489d79a7286"},
+    {file = "numpy-2.2.4-cp313-cp313t-win_amd64.whl", hash = "sha256:188dcbca89834cc2e14eb2f106c96d6d46f200fe0200310fc29089657379c58d"},
+    {file = "numpy-2.2.4-pp310-pypy310_pp73-macosx_10_15_x86_64.whl", hash = "sha256:7051ee569db5fbac144335e0f3b9c2337e0c8d5c9fee015f259a5bd70772b7e8"},
+    {file = "numpy-2.2.4-pp310-pypy310_pp73-macosx_14_0_x86_64.whl", hash = "sha256:ab2939cd5bec30a7430cbdb2287b63151b77cf9624de0532d629c9a1c59b1d5c"},
+    {file = "numpy-2.2.4-pp310-pypy310_pp73-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:d0f35b19894a9e08639fd60a1ec1978cb7f5f7f1eace62f38dd36be8aecdef4d"},
+    {file = "numpy-2.2.4-pp310-pypy310_pp73-win_amd64.whl", hash = "sha256:b4adfbbc64014976d2f91084915ca4e626fbf2057fb81af209c1a6d776d23e3d"},
+    {file = "numpy-2.2.4.tar.gz", hash = "sha256:9ba03692a45d3eef66559efe1d1096c4b9b75c0986b5dff5530c378fb8331d4f"},
 ]
 
 [[package]]
@@ -1697,7 +1836,7 @@ description = "Accelerated property cache"
 optional = false
 python-versions = ">=3.9"
 groups = ["main"]
-markers = "extra == \"graph\""
+markers = "(python_version < \"3.10\" or python_version >= \"3.12\") and extra == \"graph\""
 files = [
     {file = "propcache-0.3.0-cp310-cp310-macosx_10_9_universal2.whl", hash = "sha256:efa44f64c37cc30c9f05932c740a8b40ce359f51882c70883cc95feac842da4d"},
     {file = "propcache-0.3.0-cp310-cp310-macosx_10_9_x86_64.whl", hash = "sha256:2383a17385d9800b6eb5855c2f05ee550f803878f344f58b6e194de08b96352c"},
@@ -1930,7 +2069,7 @@ typing-extensions = ">=4.12.2"
 
 [package.extras]
 email = ["email-validator (>=2.0.0)"]
-timezone = ["tzdata"]
+timezone = ["tzdata ; python_version >= \"3.9\" and platform_system == \"Windows\""]
 
 [[package]]
 name = "pydantic-core"
@@ -2046,26 +2185,27 @@ files = [
 typing-extensions = ">=4.6.0,<4.7.0 || >4.7.0"
 
 [[package]]
-name = "pydantic-settings"
-version = "2.8.1"
-description = "Settings management using Pydantic"
+name = "pypdf"
+version = "4.3.1"
+description = "A pure-python PDF library capable of splitting, merging, cropping, and transforming PDF files"
 optional = false
-python-versions = ">=3.8"
+python-versions = ">=3.6"
 groups = ["main"]
 markers = "extra == \"graph\""
 files = [
-    {file = "pydantic_settings-2.8.1-py3-none-any.whl", hash = "sha256:81942d5ac3d905f7f3ee1a70df5dfb62d5569c12f51a5a647defc1c3d9ee2e9c"},
-    {file = "pydantic_settings-2.8.1.tar.gz", hash = "sha256:d5c663dfbe9db9d5e1c646b2e161da12f0d734d422ee56f567d0ea2cee4e8585"},
+    {file = "pypdf-4.3.1-py3-none-any.whl", hash = "sha256:64b31da97eda0771ef22edb1bfecd5deee4b72c3d1736b7df2689805076d6418"},
+    {file = "pypdf-4.3.1.tar.gz", hash = "sha256:b2f37fe9a3030aa97ca86067a56ba3f9d3565f9a791b305c7355d8392c30d91b"},
 ]
 
 [package.dependencies]
-pydantic = ">=2.7.0"
-python-dotenv = ">=0.21.0"
+typing_extensions = {version = ">=4.0", markers = "python_version < \"3.11\""}
 
 [package.extras]
-azure-key-vault = ["azure-identity (>=1.16.0)", "azure-keyvault-secrets (>=4.8.0)"]
-toml = ["tomli (>=2.0.1)"]
-yaml = ["pyyaml (>=6.0.1)"]
+crypto = ["PyCryptodome ; python_version == \"3.6\"", "cryptography ; python_version >= \"3.7\""]
+dev = ["black", "flit", "pip-tools", "pre-commit (<2.18.0)", "pytest-cov", "pytest-socket", "pytest-timeout", "pytest-xdist", "wheel"]
+docs = ["myst_parser", "sphinx", "sphinx_rtd_theme"]
+full = ["Pillow (>=8.0.0)", "PyCryptodome ; python_version == \"3.6\"", "cryptography ; python_version >= \"3.7\""]
+image = ["Pillow (>=8.0.0)"]
 
 [[package]]
 name = "pytest"
@@ -2105,22 +2245,6 @@ files = [
 [package.dependencies]
 six = ">=1.5"
 
-[[package]]
-name = "python-dotenv"
-version = "1.0.1"
-description = "Read key-value pairs from a .env file and set them as environment variables"
-optional = false
-python-versions = ">=3.8"
-groups = ["main"]
-markers = "extra == \"graph\""
-files = [
-    {file = "python-dotenv-1.0.1.tar.gz", hash = "sha256:e324ee90a023d808f1959c46bcbc04446a10ced277783dc6ee09987c37ec10ca"},
-    {file = "python_dotenv-1.0.1-py3-none-any.whl", hash = "sha256:f7b63ef50f1b690dddf550d03497b66d609393b40b564ed0d674909a68ebf16a"},
-]
-
-[package.extras]
-cli = ["click (>=5.0)"]
-
 [[package]]
 name = "pytz"
 version = "2024.2"
@@ -2245,7 +2369,7 @@ httpx = {version = ">=0.20.0", extras = ["http2"]}
 numpy = [
     {version = ">=1.21", markers = "python_version >= \"3.10\" and python_version < \"3.12\""},
     {version = ">=1.21,<2.1.0", markers = "python_version < \"3.10\""},
-    {version = ">=1.26", markers = "python_version >= \"3.12\" and python_version < \"3.13\""},
+    {version = ">=1.26", markers = "python_version == \"3.12\""},
     {version = ">=2.1.0", markers = "python_version >= \"3.13\""},
 ]
 portalocker = ">=2.7.0,<3.0.0"
@@ -2354,13 +2478,13 @@ files = [
 ]
 
 [package.extras]
-check = ["pytest-checkdocs (>=2.4)", "pytest-ruff (>=0.2.1)", "ruff (>=0.8.0)"]
-core = ["importlib_metadata (>=6)", "jaraco.collections", "jaraco.functools (>=4)", "jaraco.text (>=3.7)", "more_itertools", "more_itertools (>=8.8)", "packaging", "packaging (>=24.2)", "platformdirs (>=4.2.2)", "tomli (>=2.0.1)", "wheel (>=0.43.0)"]
+check = ["pytest-checkdocs (>=2.4)", "pytest-ruff (>=0.2.1) ; sys_platform != \"cygwin\"", "ruff (>=0.8.0) ; sys_platform != \"cygwin\""]
+core = ["importlib_metadata (>=6) ; python_version < \"3.10\"", "jaraco.collections", "jaraco.functools (>=4)", "jaraco.text (>=3.7)", "more_itertools", "more_itertools (>=8.8)", "packaging", "packaging (>=24.2)", "platformdirs (>=4.2.2)", "tomli (>=2.0.1) ; python_version < \"3.11\"", "wheel (>=0.43.0)"]
 cover = ["pytest-cov"]
 doc = ["furo", "jaraco.packaging (>=9.3)", "jaraco.tidelift (>=1.4)", "pygments-github-lexers (==0.0.5)", "pyproject-hooks (!=1.1)", "rst.linker (>=1.9)", "sphinx (>=3.5)", "sphinx-favicon", "sphinx-inline-tabs", "sphinx-lint", "sphinx-notfound-page (>=1,<2)", "sphinx-reredirects", "sphinxcontrib-towncrier", "towncrier (<24.7)"]
 enabler = ["pytest-enabler (>=2.2)"]
-test = ["build[virtualenv] (>=1.0.3)", "filelock (>=3.4.0)", "ini2toml[lite] (>=0.14)", "jaraco.develop (>=7.21)", "jaraco.envs (>=2.2)", "jaraco.path (>=3.7.2)", "jaraco.test (>=5.5)", "packaging (>=24.2)", "pip (>=19.1)", "pyproject-hooks (!=1.1)", "pytest (>=6,!=8.1.*)", "pytest-home (>=0.5)", "pytest-perf", "pytest-subprocess", "pytest-timeout", "pytest-xdist (>=3)", "tomli-w (>=1.0.0)", "virtualenv (>=13.0.0)", "wheel (>=0.44.0)"]
-type = ["importlib_metadata (>=7.0.2)", "jaraco.develop (>=7.21)", "mypy (==1.14.*)", "pytest-mypy"]
+test = ["build[virtualenv] (>=1.0.3)", "filelock (>=3.4.0)", "ini2toml[lite] (>=0.14)", "jaraco.develop (>=7.21) ; python_version >= \"3.9\" and sys_platform != \"cygwin\"", "jaraco.envs (>=2.2)", "jaraco.path (>=3.7.2)", "jaraco.test (>=5.5)", "packaging (>=24.2)", "pip (>=19.1)", "pyproject-hooks (!=1.1)", "pytest (>=6,!=8.1.*)", "pytest-home (>=0.5)", "pytest-perf ; sys_platform != \"cygwin\"", "pytest-subprocess", "pytest-timeout", "pytest-xdist (>=3)", "tomli-w (>=1.0.0)", "virtualenv (>=13.0.0)", "wheel (>=0.44.0)"]
+type = ["importlib_metadata (>=7.0.2) ; python_version < \"3.10\"", "jaraco.develop (>=7.21) ; sys_platform != \"cygwin\"", "mypy (==1.14.*)", "pytest-mypy"]
 
 [[package]]
 name = "six"
@@ -2564,6 +2688,19 @@ notebook = ["ipywidgets (>=6)"]
 slack = ["slack-sdk"]
 telegram = ["requests"]
 
+[[package]]
+name = "types-pyyaml"
+version = "6.0.12.20241230"
+description = "Typing stubs for PyYAML"
+optional = false
+python-versions = ">=3.8"
+groups = ["main"]
+markers = "extra == \"graph\""
+files = [
+    {file = "types_PyYAML-6.0.12.20241230-py3-none-any.whl", hash = "sha256:fa4d32565219b68e6dee5f67534c722e53c00d1cfc09c435ef04d7353e1e96e6"},
+    {file = "types_pyyaml-6.0.12.20241230.tar.gz", hash = "sha256:7f07622dbd34bb9c8b264fe860a17e0efcad00d50b5f27e93984909d9363498c"},
+]
+
 [[package]]
 name = "typing-extensions"
 version = "4.12.2"
@@ -2576,23 +2713,6 @@ files = [
     {file = "typing_extensions-4.12.2.tar.gz", hash = "sha256:1a7ead55c7e559dd4dee8856e3a88b41225abfe1ce8df57b7c13915fe121ffb8"},
 ]
 
-[[package]]
-name = "typing-inspect"
-version = "0.9.0"
-description = "Runtime inspection utilities for typing module."
-optional = false
-python-versions = "*"
-groups = ["main"]
-markers = "extra == \"graph\""
-files = [
-    {file = "typing_inspect-0.9.0-py3-none-any.whl", hash = "sha256:9ee6fc59062311ef8547596ab6b955e1b8aa46242d854bfc78f4f6b0eff35f9f"},
-    {file = "typing_inspect-0.9.0.tar.gz", hash = "sha256:b23fc42ff6f6ef6954e4852c1fb512cdd18dbea03134f91f856a95ccc9461f78"},
-]
-
-[package.dependencies]
-mypy-extensions = ">=0.3.0"
-typing-extensions = ">=3.7.4"
-
 [[package]]
 name = "urllib3"
 version = "2.3.0"
@@ -2606,7 +2726,7 @@ files = [
 ]
 
 [package.extras]
-brotli = ["brotli (>=1.0.9)", "brotlicffi (>=0.8.0)"]
+brotli = ["brotli (>=1.0.9) ; platform_python_implementation == \"CPython\"", "brotlicffi (>=0.8.0) ; platform_python_implementation != \"CPython\""]
 h2 = ["h2 (>=4,<5)"]
 socks = ["pysocks (>=1.5.6,!=1.5.7,<2.0)"]
 zstd = ["zstandard (>=0.18.0)"]
@@ -2618,7 +2738,7 @@ description = "Yet another URL library"
 optional = false
 python-versions = ">=3.9"
 groups = ["main"]
-markers = "extra == \"graph\""
+markers = "(python_version < \"3.10\" or python_version >= \"3.12\") and extra == \"graph\""
 files = [
     {file = "yarl-1.18.3-cp310-cp310-macosx_10_9_universal2.whl", hash = "sha256:7df647e8edd71f000a5208fe6ff8c382a1de8edfbccdbbfe649d263de07d8c34"},
     {file = "yarl-1.18.3-cp310-cp310-macosx_10_9_x86_64.whl", hash = "sha256:c69697d3adff5aa4f874b19c0e4ed65180ceed6318ec856ebc423aa5850d84f7"},
@@ -2824,9 +2944,9 @@ cffi = {version = ">=1.11", markers = "platform_python_implementation == \"PyPy\
 cffi = ["cffi (>=1.11)"]
 
 [extras]
-graph = ["langchain-community", "neo4j", "rank-bm25"]
+graph = ["langchain-neo4j", "neo4j", "rank-bm25"]
 
 [metadata]
 lock-version = "2.1"
 python-versions = ">=3.9,<4.0"
-content-hash = "3d2febfdf2d1318fc91fce5640e2f9cfb3bf964cc23a797e4f01df81dc7175da"
+content-hash = "f7bee9b294566e32c580fef1940cbc6fcb7fd6ca2e695f23c719d134af4c3204"

pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "mem0ai"
-version = "0.1.67"
+version = "0.1.70"
 description = "Long-term memory for AI Agents"
 authors = ["Mem0 <founders@mem0.ai>"]
 exclude = [
@@ -22,13 +22,14 @@ openai = "^1.33.0"
 posthog = "^3.5.0"
 pytz = "^2024.1"
 sqlalchemy = "^2.0.31"
-langchain-community = "^0.3.1"
+langchain-neo4j = "^0.4.0"
 neo4j = "^5.23.1"
 rank-bm25 = "^0.2.2"
+azure-search-documents = "^11.5.0"
 psycopg2-binary = "^2.9.10"
 
 [tool.poetry.extras]
-graph = ["langchain-community", "neo4j", "rank-bm25"]
+graph = ["langchain-neo4j", "neo4j", "rank-bm25"]
 
 [tool.poetry.group.test.dependencies]
 pytest = "^8.2.2"

tests/configs/test_prompts.py

@@ -0,0 +1,17 @@
+from mem0.configs import prompts
+
+
+def test_get_update_memory_messages():
+    retrieved_old_memory_dict = [{"id": "1", "text": "old memory 1"}]
+    response_content = ["new fact"]
+    custom_update_memory_prompt = "custom prompt determining memory update"
+    
+    ## When custom update memory prompt is provided
+    ##
+    result = prompts.get_update_memory_messages(retrieved_old_memory_dict, response_content, custom_update_memory_prompt)
+    assert result.startswith(custom_update_memory_prompt)
+    
+    ## When custom update memory prompt is not provided
+    ##
+    result = prompts.get_update_memory_messages(retrieved_old_memory_dict, response_content, None)
+    assert result.startswith(prompts.DEFAULT_UPDATE_MEMORY_PROMPT)

tests/llms/test_azure_openai.py

@@ -20,8 +20,10 @@ def mock_openai_client():
         yield mock_client
 
 
-def test_generate_response_without_tools(mock_openai_client):
-    config = BaseLlmConfig(model=MODEL, temperature=TEMPERATURE, max_tokens=MAX_TOKENS, top_p=TOP_P)
+def test_generate_response(mock_openai_client):
+    config = BaseLlmConfig(
+        model=MODEL, temperature=TEMPERATURE, max_tokens=MAX_TOKENS, top_p=TOP_P
+    )
     llm = AzureOpenAILLM(config)
     messages = [
         {"role": "system", "content": "You are a helpful assistant."},
@@ -29,67 +31,21 @@ def test_generate_response_without_tools(mock_openai_client):
     ]
 
     mock_response = Mock()
-    mock_response.choices = [Mock(message=Mock(content="I'm doing well, thank you for asking!"))]
+    mock_response.choices = [
+        Mock(message=Mock(content="I'm doing well, thank you for asking!"))
+    ]
     mock_openai_client.chat.completions.create.return_value = mock_response
 
     response = llm.generate_response(messages)
 
-    mock_openai_client.chat.completions.create.assert_called_once_with(
-        model=MODEL, messages=messages, temperature=TEMPERATURE, max_tokens=MAX_TOKENS, top_p=TOP_P
-    )
-    assert response == "I'm doing well, thank you for asking!"
-
-
-def test_generate_response_with_tools(mock_openai_client):
-    config = BaseLlmConfig(model=MODEL, temperature=TEMPERATURE, max_tokens=MAX_TOKENS, top_p=TOP_P)
-    llm = AzureOpenAILLM(config)
-    messages = [
-        {"role": "system", "content": "You are a helpful assistant."},
-        {"role": "user", "content": "Add a new memory: Today is a sunny day."},
-    ]
-    tools = [
-        {
-            "type": "function",
-            "function": {
-                "name": "add_memory",
-                "description": "Add a memory",
-                "parameters": {
-                    "type": "object",
-                    "properties": {"data": {"type": "string", "description": "Data to add to memory"}},
-                    "required": ["data"],
-                },
-            },
-        }
-    ]
-
-    mock_response = Mock()
-    mock_message = Mock()
-    mock_message.content = "I've added the memory for you."
-
-    mock_tool_call = Mock()
-    mock_tool_call.function.name = "add_memory"
-    mock_tool_call.function.arguments = '{"data": "Today is a sunny day."}'
-
-    mock_message.tool_calls = [mock_tool_call]
-    mock_response.choices = [Mock(message=mock_message)]
-    mock_openai_client.chat.completions.create.return_value = mock_response
-
-    response = llm.generate_response(messages, tools=tools)
-
     mock_openai_client.chat.completions.create.assert_called_once_with(
         model=MODEL,
         messages=messages,
         temperature=TEMPERATURE,
         max_tokens=MAX_TOKENS,
         top_p=TOP_P,
-        tools=tools,
-        tool_choice="auto",
     )
-
-    assert response["content"] == "I've added the memory for you."
-    assert len(response["tool_calls"]) == 1
-    assert response["tool_calls"][0]["name"] == "add_memory"
-    assert response["tool_calls"][0]["arguments"] == {"data": "Today is a sunny day."}
+    assert response == "I'm doing well, thank you for asking!"
 
 
 @pytest.mark.parametrize(
@@ -128,4 +84,6 @@ def test_generate_with_http_proxies(default_headers):
             api_version=None,
             default_headers=default_headers,
         )
-        mock_http_client.assert_called_once_with(proxies="http://testproxy.mem0.net:8000")
+        mock_http_client.assert_called_once_with(
+            proxies="http://testproxy.mem0.net:8000"
+        )

tests/llms/test_deepseek.py

@@ -16,33 +16,47 @@ def mock_deepseek_client():
 
 def test_deepseek_llm_base_url():
     # case1: default config with deepseek official base url
-    config = BaseLlmConfig(model="deepseek-chat", temperature=0.7, max_tokens=100, top_p=1.0, api_key="api_key")
+    config = BaseLlmConfig(
+        model="deepseek-chat",
+        temperature=0.7,
+        max_tokens=100,
+        top_p=1.0,
+        api_key="api_key",
+    )
     llm = DeepSeekLLM(config)
     assert str(llm.client.base_url) == "https://api.deepseek.com"
 
     # case2: with env variable DEEPSEEK_API_BASE
     provider_base_url = "https://api.provider.com/v1/"
     os.environ["DEEPSEEK_API_BASE"] = provider_base_url
-    config = BaseLlmConfig(model="deepseek-chat", temperature=0.7, max_tokens=100, top_p=1.0, api_key="api_key")
+    config = BaseLlmConfig(
+        model="deepseek-chat",
+        temperature=0.7,
+        max_tokens=100,
+        top_p=1.0,
+        api_key="api_key",
+    )
     llm = DeepSeekLLM(config)
     assert str(llm.client.base_url) == provider_base_url
 
     # case3: with config.deepseek_base_url
     config_base_url = "https://api.config.com/v1/"
     config = BaseLlmConfig(
-        model="deepseek-chat", 
-        temperature=0.7, 
-        max_tokens=100, 
-        top_p=1.0, 
-        api_key="api_key", 
-        deepseek_base_url=config_base_url
+        model="deepseek-chat",
+        temperature=0.7,
+        max_tokens=100,
+        top_p=1.0,
+        api_key="api_key",
+        deepseek_base_url=config_base_url,
     )
     llm = DeepSeekLLM(config)
     assert str(llm.client.base_url) == config_base_url
 
 
-def test_generate_response_without_tools(mock_deepseek_client):
-    config = BaseLlmConfig(model="deepseek-chat", temperature=0.7, max_tokens=100, top_p=1.0)
+def test_generate_response(mock_deepseek_client):
+    config = BaseLlmConfig(
+        model="deepseek-chat", temperature=0.7, max_tokens=100, top_p=1.0
+    )
     llm = DeepSeekLLM(config)
     messages = [
         {"role": "system", "content": "You are a helpful assistant."},
@@ -50,64 +64,18 @@ def test_generate_response_without_tools(mock_deepseek_client):
     ]
 
     mock_response = Mock()
-    mock_response.choices = [Mock(message=Mock(content="I'm doing well, thank you for asking!"))]
+    mock_response.choices = [
+        Mock(message=Mock(content="I'm doing well, thank you for asking!"))
+    ]
     mock_deepseek_client.chat.completions.create.return_value = mock_response
 
     response = llm.generate_response(messages)
 
     mock_deepseek_client.chat.completions.create.assert_called_once_with(
-        model="deepseek-chat", messages=messages, temperature=0.7, max_tokens=100, top_p=1.0
+        model="deepseek-chat",
+        messages=messages,
+        temperature=0.7,
+        max_tokens=100,
+        top_p=1.0,
     )
     assert response == "I'm doing well, thank you for asking!"
-
-
-def test_generate_response_with_tools(mock_deepseek_client):
-    config = BaseLlmConfig(model="deepseek-chat", temperature=0.7, max_tokens=100, top_p=1.0)
-    llm = DeepSeekLLM(config)
-    messages = [
-        {"role": "system", "content": "You are a helpful assistant."},
-        {"role": "user", "content": "Add a new memory: Today is a sunny day."},
-    ]
-    tools = [
-        {
-            "type": "function",
-            "function": {
-                "name": "add_memory",
-                "description": "Add a memory",
-                "parameters": {
-                    "type": "object",
-                    "properties": {"data": {"type": "string", "description": "Data to add to memory"}},
-                    "required": ["data"],
-                },
-            },
-        }
-    ]
-
-    mock_response = Mock()
-    mock_message = Mock()
-    mock_message.content = "I've added the memory for you."
-
-    mock_tool_call = Mock()
-    mock_tool_call.function.name = "add_memory"
-    mock_tool_call.function.arguments = '{"data": "Today is a sunny day."}'
-
-    mock_message.tool_calls = [mock_tool_call]
-    mock_response.choices = [Mock(message=mock_message)]
-    mock_deepseek_client.chat.completions.create.return_value = mock_response
-
-    response = llm.generate_response(messages, tools=tools)
-
-    mock_deepseek_client.chat.completions.create.assert_called_once_with(
-        model="deepseek-chat", 
-        messages=messages, 
-        temperature=0.7, 
-        max_tokens=100, 
-        top_p=1.0, 
-        tools=tools, 
-        tool_choice="auto"
-    )
-
-    assert response["content"] == "I've added the memory for you."
-    assert len(response["tool_calls"]) == 1
-    assert response["tool_calls"][0]["name"] == "add_memory"
-    assert response["tool_calls"][0]["arguments"] == {"data": "Today is a sunny day."}

tests/llms/test_gemini_llm.py

@@ -17,7 +17,9 @@ def mock_gemini_client():
 
 
 def test_generate_response_without_tools(mock_gemini_client: Mock):
-    config = BaseLlmConfig(model="gemini-1.5-flash-latest", temperature=0.7, max_tokens=100, top_p=1.0)
+    config = BaseLlmConfig(
+        model="gemini-1.5-flash-latest", temperature=0.7, max_tokens=100, top_p=1.0
+    )
     llm = GeminiLLM(config)
     messages = [
         {"role": "system", "content": "You are a helpful assistant."},
@@ -34,86 +36,14 @@ def test_generate_response_without_tools(mock_gemini_client: Mock):
 
     mock_gemini_client.generate_content.assert_called_once_with(
         contents=[
-            {"parts": "THIS IS A SYSTEM PROMPT. YOU MUST OBEY THIS: You are a helpful assistant.", "role": "user"},
+            {
+                "parts": "THIS IS A SYSTEM PROMPT. YOU MUST OBEY THIS: You are a helpful assistant.",
+                "role": "user",
+            },
             {"parts": "Hello, how are you?", "role": "user"},
         ],
-        generation_config=GenerationConfig(temperature=0.7, max_output_tokens=100, top_p=1.0),
-        tools=None,
-        tool_config=content_types.to_tool_config(
-            {"function_calling_config": {"mode": "auto", "allowed_function_names": None}}
+        generation_config=GenerationConfig(
+            temperature=0.7, max_output_tokens=100, top_p=1.0
         ),
     )
     assert response == "I'm doing well, thank you for asking!"
-
-
-def test_generate_response_with_tools(mock_gemini_client: Mock):
-    config = BaseLlmConfig(model="gemini-1.5-flash-latest", temperature=0.7, max_tokens=100, top_p=1.0)
-    llm = GeminiLLM(config)
-    messages = [
-        {"role": "system", "content": "You are a helpful assistant."},
-        {"role": "user", "content": "Add a new memory: Today is a sunny day."},
-    ]
-    tools = [
-        {
-            "type": "function",
-            "function": {
-                "name": "add_memory",
-                "description": "Add a memory",
-                "parameters": {
-                    "type": "object",
-                    "properties": {"data": {"type": "string", "description": "Data to add to memory"}},
-                    "required": ["data"],
-                },
-            },
-        }
-    ]
-
-    mock_tool_call = Mock()
-    mock_tool_call.name = "add_memory"
-    mock_tool_call.args = {"data": "Today is a sunny day."}
-
-    mock_part = Mock()
-    mock_part.function_call = mock_tool_call
-    mock_part.text = "I've added the memory for you."
-
-    mock_content = Mock()
-    mock_content.parts = [mock_part]
-
-    mock_message = Mock()
-    mock_message.content = mock_content
-
-    mock_response = Mock(candidates=[mock_message])
-    mock_gemini_client.generate_content.return_value = mock_response
-
-    response = llm.generate_response(messages, tools=tools)
-
-    mock_gemini_client.generate_content.assert_called_once_with(
-        contents=[
-            {"parts": "THIS IS A SYSTEM PROMPT. YOU MUST OBEY THIS: You are a helpful assistant.", "role": "user"},
-            {"parts": "Add a new memory: Today is a sunny day.", "role": "user"},
-        ],
-        generation_config=GenerationConfig(temperature=0.7, max_output_tokens=100, top_p=1.0),
-        tools=[
-            {
-                "function_declarations": [
-                    {
-                        "name": "add_memory",
-                        "description": "Add a memory",
-                        "parameters": {
-                            "type": "object",
-                            "properties": {"data": {"type": "string", "description": "Data to add to memory"}},
-                            "required": ["data"],
-                        },
-                    }
-                ]
-            }
-        ],
-        tool_config=content_types.to_tool_config(
-            {"function_calling_config": {"mode": "auto", "allowed_function_names": None}}
-        ),
-    )
-
-    assert response["content"] == "I've added the memory for you."
-    assert len(response["tool_calls"]) == 1
-    assert response["tool_calls"][0]["name"] == "add_memory"
-    assert response["tool_calls"][0]["arguments"] == {"data": "Today is a sunny day."}

tests/llms/test_groq.py

@@ -14,8 +14,10 @@ def mock_groq_client():
         yield mock_client
 
 
-def test_generate_response_without_tools(mock_groq_client):
-    config = BaseLlmConfig(model="llama3-70b-8192", temperature=0.7, max_tokens=100, top_p=1.0)
+def test_generate_response(mock_groq_client):
+    config = BaseLlmConfig(
+        model="llama3-70b-8192", temperature=0.7, max_tokens=100, top_p=1.0
+    )
     llm = GroqLLM(config)
     messages = [
         {"role": "system", "content": "You are a helpful assistant."},
@@ -23,64 +25,18 @@ def test_generate_response_without_tools(mock_groq_client):
     ]
 
     mock_response = Mock()
-    mock_response.choices = [Mock(message=Mock(content="I'm doing well, thank you for asking!"))]
+    mock_response.choices = [
+        Mock(message=Mock(content="I'm doing well, thank you for asking!"))
+    ]
     mock_groq_client.chat.completions.create.return_value = mock_response
 
     response = llm.generate_response(messages)
 
-    mock_groq_client.chat.completions.create.assert_called_once_with(
-        model="llama3-70b-8192", messages=messages, temperature=0.7, max_tokens=100, top_p=1.0
-    )
-    assert response == "I'm doing well, thank you for asking!"
-
-
-def test_generate_response_with_tools(mock_groq_client):
-    config = BaseLlmConfig(model="llama3-70b-8192", temperature=0.7, max_tokens=100, top_p=1.0)
-    llm = GroqLLM(config)
-    messages = [
-        {"role": "system", "content": "You are a helpful assistant."},
-        {"role": "user", "content": "Add a new memory: Today is a sunny day."},
-    ]
-    tools = [
-        {
-            "type": "function",
-            "function": {
-                "name": "add_memory",
-                "description": "Add a memory",
-                "parameters": {
-                    "type": "object",
-                    "properties": {"data": {"type": "string", "description": "Data to add to memory"}},
-                    "required": ["data"],
-                },
-            },
-        }
-    ]
-
-    mock_response = Mock()
-    mock_message = Mock()
-    mock_message.content = "I've added the memory for you."
-
-    mock_tool_call = Mock()
-    mock_tool_call.function.name = "add_memory"
-    mock_tool_call.function.arguments = '{"data": "Today is a sunny day."}'
-
-    mock_message.tool_calls = [mock_tool_call]
-    mock_response.choices = [Mock(message=mock_message)]
-    mock_groq_client.chat.completions.create.return_value = mock_response
-
-    response = llm.generate_response(messages, tools=tools)
-
     mock_groq_client.chat.completions.create.assert_called_once_with(
         model="llama3-70b-8192",
         messages=messages,
         temperature=0.7,
         max_tokens=100,
         top_p=1.0,
-        tools=tools,
-        tool_choice="auto",
     )
-
-    assert response["content"] == "I've added the memory for you."
-    assert len(response["tool_calls"]) == 1
-    assert response["tool_calls"][0]["name"] == "add_memory"
-    assert response["tool_calls"][0]["arguments"] == {"data": "Today is a sunny day."}
+    assert response == "I'm doing well, thank you for asking!"

tests/llms/test_litellm.py

@@ -13,17 +13,22 @@ def mock_litellm():
 
 
 def test_generate_response_with_unsupported_model(mock_litellm):
-    config = BaseLlmConfig(model="unsupported-model", temperature=0.7, max_tokens=100, top_p=1)
+    config = BaseLlmConfig(
+        model="unsupported-model", temperature=0.7, max_tokens=100, top_p=1
+    )
     llm = litellm.LiteLLM(config)
     messages = [{"role": "user", "content": "Hello"}]
 
     mock_litellm.supports_function_calling.return_value = False
 
-    with pytest.raises(ValueError, match="Model 'unsupported-model' in litellm does not support function calling."):
+    with pytest.raises(
+        ValueError,
+        match="Model 'unsupported-model' in LiteLLM does not support function calling.",
+    ):
         llm.generate_response(messages)
 
 
-def test_generate_response_without_tools(mock_litellm):
+def test_generate_response(mock_litellm):
     config = BaseLlmConfig(model="gpt-4o", temperature=0.7, max_tokens=100, top_p=1)
     llm = litellm.LiteLLM(config)
     messages = [
@@ -32,7 +37,9 @@ def test_generate_response_without_tools(mock_litellm):
     ]
 
     mock_response = Mock()
-    mock_response.choices = [Mock(message=Mock(content="I'm doing well, thank you for asking!"))]
+    mock_response.choices = [
+        Mock(message=Mock(content="I'm doing well, thank you for asking!"))
+    ]
     mock_litellm.completion.return_value = mock_response
     mock_litellm.supports_function_calling.return_value = True
 
@@ -42,50 +49,3 @@ def test_generate_response_without_tools(mock_litellm):
         model="gpt-4o", messages=messages, temperature=0.7, max_tokens=100, top_p=1.0
     )
     assert response == "I'm doing well, thank you for asking!"
-
-
-def test_generate_response_with_tools(mock_litellm):
-    config = BaseLlmConfig(model="gpt-4o", temperature=0.7, max_tokens=100, top_p=1)
-    llm = litellm.LiteLLM(config)
-    messages = [
-        {"role": "system", "content": "You are a helpful assistant."},
-        {"role": "user", "content": "Add a new memory: Today is a sunny day."},
-    ]
-    tools = [
-        {
-            "type": "function",
-            "function": {
-                "name": "add_memory",
-                "description": "Add a memory",
-                "parameters": {
-                    "type": "object",
-                    "properties": {"data": {"type": "string", "description": "Data to add to memory"}},
-                    "required": ["data"],
-                },
-            },
-        }
-    ]
-
-    mock_response = Mock()
-    mock_message = Mock()
-    mock_message.content = "I've added the memory for you."
-
-    mock_tool_call = Mock()
-    mock_tool_call.function.name = "add_memory"
-    mock_tool_call.function.arguments = '{"data": "Today is a sunny day."}'
-
-    mock_message.tool_calls = [mock_tool_call]
-    mock_response.choices = [Mock(message=mock_message)]
-    mock_litellm.completion.return_value = mock_response
-    mock_litellm.supports_function_calling.return_value = True
-
-    response = llm.generate_response(messages, tools=tools)
-
-    mock_litellm.completion.assert_called_once_with(
-        model="gpt-4o", messages=messages, temperature=0.7, max_tokens=100, top_p=1, tools=tools, tool_choice="auto"
-    )
-
-    assert response["content"] == "I've added the memory for you."
-    assert len(response["tool_calls"]) == 1
-    assert response["tool_calls"][0]["name"] == "add_memory"
-    assert response["tool_calls"][0]["arguments"] == {"data": "Today is a sunny day."}

tests/llms/test_openai.py

@@ -16,7 +16,9 @@ def mock_openai_client():
 
 def test_openai_llm_base_url():
     # case1: default config: with openai official base url
-    config = BaseLlmConfig(model="gpt-4o", temperature=0.7, max_tokens=100, top_p=1.0, api_key="api_key")
+    config = BaseLlmConfig(
+        model="gpt-4o", temperature=0.7, max_tokens=100, top_p=1.0, api_key="api_key"
+    )
     llm = OpenAILLM(config)
     # Note: openai client will parse the raw base_url into a URL object, which will have a trailing slash
     assert str(llm.client.base_url) == "https://api.openai.com/v1/"
@@ -24,7 +26,9 @@ def test_openai_llm_base_url():
     # case2: with env variable OPENAI_API_BASE
     provider_base_url = "https://api.provider.com/v1"
     os.environ["OPENAI_API_BASE"] = provider_base_url
-    config = BaseLlmConfig(model="gpt-4o", temperature=0.7, max_tokens=100, top_p=1.0, api_key="api_key")
+    config = BaseLlmConfig(
+        model="gpt-4o", temperature=0.7, max_tokens=100, top_p=1.0, api_key="api_key"
+    )
     llm = OpenAILLM(config)
     # Note: openai client will parse the raw base_url into a URL object, which will have a trailing slash
     assert str(llm.client.base_url) == provider_base_url + "/"
@@ -32,14 +36,19 @@ def test_openai_llm_base_url():
     # case3: with config.openai_base_url
     config_base_url = "https://api.config.com/v1"
     config = BaseLlmConfig(
-        model="gpt-4o", temperature=0.7, max_tokens=100, top_p=1.0, api_key="api_key", openai_base_url=config_base_url
+        model="gpt-4o",
+        temperature=0.7,
+        max_tokens=100,
+        top_p=1.0,
+        api_key="api_key",
+        openai_base_url=config_base_url,
     )
     llm = OpenAILLM(config)
     # Note: openai client will parse the raw base_url into a URL object, which will have a trailing slash
     assert str(llm.client.base_url) == config_base_url + "/"
 
 
-def test_generate_response_without_tools(mock_openai_client):
+def test_generate_response(mock_openai_client):
     config = BaseLlmConfig(model="gpt-4o", temperature=0.7, max_tokens=100, top_p=1.0)
     llm = OpenAILLM(config)
     messages = [
@@ -48,7 +57,9 @@ def test_generate_response_without_tools(mock_openai_client):
     ]
 
     mock_response = Mock()
-    mock_response.choices = [Mock(message=Mock(content="I'm doing well, thank you for asking!"))]
+    mock_response.choices = [
+        Mock(message=Mock(content="I'm doing well, thank you for asking!"))
+    ]
     mock_openai_client.chat.completions.create.return_value = mock_response
 
     response = llm.generate_response(messages)
@@ -57,49 +68,3 @@ def test_generate_response_without_tools(mock_openai_client):
         model="gpt-4o", messages=messages, temperature=0.7, max_tokens=100, top_p=1.0
     )
     assert response == "I'm doing well, thank you for asking!"
-
-
-def test_generate_response_with_tools(mock_openai_client):
-    config = BaseLlmConfig(model="gpt-4o", temperature=0.7, max_tokens=100, top_p=1.0)
-    llm = OpenAILLM(config)
-    messages = [
-        {"role": "system", "content": "You are a helpful assistant."},
-        {"role": "user", "content": "Add a new memory: Today is a sunny day."},
-    ]
-    tools = [
-        {
-            "type": "function",
-            "function": {
-                "name": "add_memory",
-                "description": "Add a memory",
-                "parameters": {
-                    "type": "object",
-                    "properties": {"data": {"type": "string", "description": "Data to add to memory"}},
-                    "required": ["data"],
-                },
-            },
-        }
-    ]
-
-    mock_response = Mock()
-    mock_message = Mock()
-    mock_message.content = "I've added the memory for you."
-
-    mock_tool_call = Mock()
-    mock_tool_call.function.name = "add_memory"
-    mock_tool_call.function.arguments = '{"data": "Today is a sunny day."}'
-
-    mock_message.tool_calls = [mock_tool_call]
-    mock_response.choices = [Mock(message=mock_message)]
-    mock_openai_client.chat.completions.create.return_value = mock_response
-
-    response = llm.generate_response(messages, tools=tools)
-
-    mock_openai_client.chat.completions.create.assert_called_once_with(
-        model="gpt-4o", messages=messages, temperature=0.7, max_tokens=100, top_p=1.0, tools=tools, tool_choice="auto"
-    )
-
-    assert response["content"] == "I've added the memory for you."
-    assert len(response["tool_calls"]) == 1
-    assert response["tool_calls"][0]["name"] == "add_memory"
-    assert response["tool_calls"][0]["arguments"] == {"data": "Today is a sunny day."}

tests/llms/test_together.py

@@ -14,8 +14,13 @@ def mock_together_client():
         yield mock_client
 
 
-def test_generate_response_without_tools(mock_together_client):
-    config = BaseLlmConfig(model="mistralai/Mixtral-8x7B-Instruct-v0.1", temperature=0.7, max_tokens=100, top_p=1.0)
+def test_generate_response(mock_together_client):
+    config = BaseLlmConfig(
+        model="mistralai/Mixtral-8x7B-Instruct-v0.1",
+        temperature=0.7,
+        max_tokens=100,
+        top_p=1.0,
+    )
     llm = TogetherLLM(config)
     messages = [
         {"role": "system", "content": "You are a helpful assistant."},
@@ -23,64 +28,18 @@ def test_generate_response_without_tools(mock_together_client):
     ]
 
     mock_response = Mock()
-    mock_response.choices = [Mock(message=Mock(content="I'm doing well, thank you for asking!"))]
+    mock_response.choices = [
+        Mock(message=Mock(content="I'm doing well, thank you for asking!"))
+    ]
     mock_together_client.chat.completions.create.return_value = mock_response
 
     response = llm.generate_response(messages)
 
-    mock_together_client.chat.completions.create.assert_called_once_with(
-        model="mistralai/Mixtral-8x7B-Instruct-v0.1", messages=messages, temperature=0.7, max_tokens=100, top_p=1.0
-    )
-    assert response == "I'm doing well, thank you for asking!"
-
-
-def test_generate_response_with_tools(mock_together_client):
-    config = BaseLlmConfig(model="mistralai/Mixtral-8x7B-Instruct-v0.1", temperature=0.7, max_tokens=100, top_p=1.0)
-    llm = TogetherLLM(config)
-    messages = [
-        {"role": "system", "content": "You are a helpful assistant."},
-        {"role": "user", "content": "Add a new memory: Today is a sunny day."},
-    ]
-    tools = [
-        {
-            "type": "function",
-            "function": {
-                "name": "add_memory",
-                "description": "Add a memory",
-                "parameters": {
-                    "type": "object",
-                    "properties": {"data": {"type": "string", "description": "Data to add to memory"}},
-                    "required": ["data"],
-                },
-            },
-        }
-    ]
-
-    mock_response = Mock()
-    mock_message = Mock()
-    mock_message.content = "I've added the memory for you."
-
-    mock_tool_call = Mock()
-    mock_tool_call.function.name = "add_memory"
-    mock_tool_call.function.arguments = '{"data": "Today is a sunny day."}'
-
-    mock_message.tool_calls = [mock_tool_call]
-    mock_response.choices = [Mock(message=mock_message)]
-    mock_together_client.chat.completions.create.return_value = mock_response
-
-    response = llm.generate_response(messages, tools=tools)
-
     mock_together_client.chat.completions.create.assert_called_once_with(
         model="mistralai/Mixtral-8x7B-Instruct-v0.1",
         messages=messages,
         temperature=0.7,
         max_tokens=100,
         top_p=1.0,
-        tools=tools,
-        tool_choice="auto",
     )
-
-    assert response["content"] == "I've added the memory for you."
-    assert len(response["tool_calls"]) == 1
-    assert response["tool_calls"][0]["name"] == "add_memory"
-    assert response["tool_calls"][0]["arguments"] == {"data": "Today is a sunny day."}
+    assert response == "I'm doing well, thank you for asking!"

tests/test_main.py

@@ -30,6 +30,25 @@ def memory_instance():
         config = MemoryConfig(version="v1.1")
         config.graph_store.config = {"some_config": "value"}
         return Memory(config)
+    
+@pytest.fixture
+def memory_custom_instance():
+    with patch("mem0.utils.factory.EmbedderFactory") as mock_embedder, patch(
+        "mem0.utils.factory.VectorStoreFactory"
+    ) as mock_vector_store, patch("mem0.utils.factory.LlmFactory") as mock_llm, patch(
+        "mem0.memory.telemetry.capture_event"
+    ), patch("mem0.memory.graph_memory.MemoryGraph"):
+        mock_embedder.create.return_value = Mock()
+        mock_vector_store.create.return_value = Mock()
+        mock_llm.create.return_value = Mock()
+
+        config = MemoryConfig(
+            version="v1.1",
+            custom_fact_extraction_prompt="custom prompt extracting memory",
+            custom_update_memory_prompt="custom prompt determining memory update"
+        )
+        config.graph_store.config = {"some_config": "value"}
+        return Memory(config)
 
 
 @pytest.mark.parametrize("version, enable_graph", [("v1.0", False), ("v1.1", True)])
@@ -51,7 +70,7 @@ def test_add(memory_instance, version, enable_graph):
         assert result["results"] == [{"memory": "Test memory", "event": "ADD"}]
 
     memory_instance._add_to_vector_store.assert_called_once_with(
-        [{"role": "user", "content": "Test message"}], {"user_id": "test_user"}, {"user_id": "test_user"}
+        [{"role": "user", "content": "Test message"}], {"user_id": "test_user"}, {"user_id": "test_user"}, True
     )
 
     # Remove the conditional assertion for _add_to_graph
@@ -239,3 +258,33 @@ def test_get_all(memory_instance, version, enable_graph, expected_result):
         memory_instance.graph.get_all.assert_called_once_with({"user_id": "test_user"}, 100)
     else:
         memory_instance.graph.get_all.assert_not_called()
+        
+
+def test_custom_prompts(memory_custom_instance):
+    messages = [{"role": "user", "content": "Test message"}]
+    memory_custom_instance.llm.generate_response = Mock()
+    
+    with patch("mem0.memory.main.parse_messages", return_value="Test message") as mock_parse_messages:
+        with patch("mem0.memory.main.get_update_memory_messages", return_value="custom update memory prompt") as mock_get_update_memory_messages:
+            memory_custom_instance.add(messages=messages, user_id="test_user")
+            
+            ## custom prompt
+            ##
+            mock_parse_messages.assert_called_once_with(messages)
+            
+            memory_custom_instance.llm.generate_response.assert_any_call(
+                messages=[
+                    {"role": "system", "content": memory_custom_instance.config.custom_fact_extraction_prompt},
+                    {"role": "user", "content": f"Input:\n{mock_parse_messages.return_value}"},
+                ],
+                response_format={"type": "json_object"},
+            )
+            
+            ## custom update memory prompt
+            ##
+            mock_get_update_memory_messages.assert_called_once_with([],[],memory_custom_instance.config.custom_update_memory_prompt)
+            
+            memory_custom_instance.llm.generate_response.assert_any_call(
+                messages=[{"role": "user", "content": mock_get_update_memory_messages.return_value}],
+                response_format={"type": "json_object"},
+            )

tests/vector_stores/test_azure_ai_search.py

@@ -0,0 +1,542 @@
+import json
+from unittest.mock import Mock, patch, MagicMock, call
+import pytest
+from azure.core.exceptions import ResourceNotFoundError, HttpResponseError
+
+# Import the AzureAISearch class and related models
+from mem0.vector_stores.azure_ai_search import AzureAISearch, OutputData
+from mem0.configs.vector_stores.azure_ai_search import AzureAISearchConfig
+
+
+# Fixture to patch SearchClient and SearchIndexClient and create an instance of AzureAISearch.
+@pytest.fixture
+def mock_clients():
+    with patch("mem0.vector_stores.azure_ai_search.SearchClient") as MockSearchClient, \
+         patch("mem0.vector_stores.azure_ai_search.SearchIndexClient") as MockIndexClient, \
+         patch("mem0.vector_stores.azure_ai_search.AzureKeyCredential") as MockAzureKeyCredential:
+        # Create mocked instances for search and index clients.
+        mock_search_client = MockSearchClient.return_value
+        mock_index_client = MockIndexClient.return_value
+        
+        # Mock the client._client._config.user_agent_policy.add_user_agent
+        mock_search_client._client = MagicMock()
+        mock_search_client._client._config.user_agent_policy.add_user_agent = Mock()
+        mock_index_client._client = MagicMock()
+        mock_index_client._client._config.user_agent_policy.add_user_agent = Mock()
+
+        # Stub required methods on search_client.
+        mock_search_client.upload_documents = Mock()
+        mock_search_client.upload_documents.return_value = [{"status": True, "id": "doc1"}]
+        mock_search_client.search = Mock()
+        mock_search_client.delete_documents = Mock()
+        mock_search_client.delete_documents.return_value = [{"status": True, "id": "doc1"}]
+        mock_search_client.merge_or_upload_documents = Mock()
+        mock_search_client.merge_or_upload_documents.return_value = [{"status": True, "id": "doc1"}]
+        mock_search_client.get_document = Mock()
+        mock_search_client.close = Mock()
+
+        # Stub required methods on index_client.
+        mock_index_client.create_or_update_index = Mock()
+        mock_index_client.list_indexes = Mock()
+        mock_index_client.list_index_names = Mock(return_value=["test-index"])
+        mock_index_client.delete_index = Mock()
+        # For col_info() we assume get_index returns an object with name and fields attributes.
+        fake_index = Mock()
+        fake_index.name = "test-index"
+        fake_index.fields = ["id", "vector", "payload", "user_id", "run_id", "agent_id"]
+        mock_index_client.get_index = Mock(return_value=fake_index)
+        mock_index_client.close = Mock()
+
+        yield mock_search_client, mock_index_client, MockAzureKeyCredential
+
+
+@pytest.fixture
+def azure_ai_search_instance(mock_clients):
+    mock_search_client, mock_index_client, _ = mock_clients
+    # Create an instance with dummy parameters.
+    instance = AzureAISearch(
+        service_name="test-service",
+        collection_name="test-index",
+        api_key="test-api-key",
+        embedding_model_dims=3,
+        compression_type="binary",  # testing binary quantization option
+        use_float16=True
+    )
+    # Return instance and clients for verification.
+    return instance, mock_search_client, mock_index_client
+
+
+# --- Tests for AzureAISearchConfig ---
+
+def test_config_validation_valid():
+    """Test valid configurations are accepted."""
+    # Test minimal configuration
+    config = AzureAISearchConfig(
+        service_name="test-service",
+        api_key="test-api-key",
+        embedding_model_dims=768
+    )
+    assert config.collection_name == "mem0"  # Default value
+    assert config.service_name == "test-service"
+    assert config.api_key == "test-api-key"
+    assert config.embedding_model_dims == 768
+    assert config.compression_type is None
+    assert config.use_float16 is False
+    
+    # Test with all optional parameters
+    config = AzureAISearchConfig(
+        collection_name="custom-index",
+        service_name="test-service",
+        api_key="test-api-key",
+        embedding_model_dims=1536,
+        compression_type="scalar",
+        use_float16=True
+    )
+    assert config.collection_name == "custom-index"
+    assert config.compression_type == "scalar"
+    assert config.use_float16 is True
+
+
+def test_config_validation_invalid_compression_type():
+    """Test that invalid compression types are rejected."""
+    with pytest.raises(ValueError) as exc_info:
+        AzureAISearchConfig(
+            service_name="test-service",
+            api_key="test-api-key",
+            embedding_model_dims=768,
+            compression_type="invalid-type"  # Not a valid option
+        )
+    assert "Invalid compression_type" in str(exc_info.value)
+
+
+def test_config_validation_deprecated_use_compression():
+    """Test that using the deprecated use_compression parameter raises an error."""
+    with pytest.raises(ValueError) as exc_info:
+        AzureAISearchConfig(
+            service_name="test-service",
+            api_key="test-api-key",
+            embedding_model_dims=768,
+            use_compression=True  # Deprecated parameter
+        )
+    # Fix: Use a partial string match instead of exact match
+    assert "use_compression" in str(exc_info.value)
+    assert "no longer supported" in str(exc_info.value)
+
+
+def test_config_validation_extra_fields():
+    """Test that extra fields are rejected."""
+    with pytest.raises(ValueError) as exc_info:
+        AzureAISearchConfig(
+            service_name="test-service",
+            api_key="test-api-key",
+            embedding_model_dims=768,
+            unknown_parameter="value"  # Extra field
+        )
+    assert "Extra fields not allowed" in str(exc_info.value)
+    assert "unknown_parameter" in str(exc_info.value)
+
+
+# --- Tests for AzureAISearch initialization ---
+
+def test_initialization(mock_clients):
+    """Test AzureAISearch initialization with different parameters."""
+    mock_search_client, mock_index_client, mock_azure_key_credential = mock_clients
+    
+    # Test with minimal parameters
+    instance = AzureAISearch(
+        service_name="test-service",
+        collection_name="test-index",
+        api_key="test-api-key",
+        embedding_model_dims=768
+    )
+    
+    # Verify initialization parameters
+    assert instance.index_name == "test-index"
+    assert instance.collection_name == "test-index"
+    assert instance.embedding_model_dims == 768
+    assert instance.compression_type == "none"  # Default when None is passed
+    assert instance.use_float16 is False
+    
+    # Verify client creation
+    mock_azure_key_credential.assert_called_with("test-api-key")
+    assert "mem0" in mock_search_client._client._config.user_agent_policy.add_user_agent.call_args[0]
+    assert "mem0" in mock_index_client._client._config.user_agent_policy.add_user_agent.call_args[0]
+    
+    # Verify index creation was called
+    mock_index_client.create_or_update_index.assert_called_once()
+
+
+def test_initialization_with_compression_types(mock_clients):
+    """Test initialization with different compression types."""
+    mock_search_client, mock_index_client, _ = mock_clients
+    
+    # Test with scalar compression
+    instance = AzureAISearch(
+        service_name="test-service",
+        collection_name="scalar-index",
+        api_key="test-api-key",
+        embedding_model_dims=768,
+        compression_type="scalar"
+    )
+    assert instance.compression_type == "scalar"
+    
+    # Capture the index creation call
+    args, _ = mock_index_client.create_or_update_index.call_args_list[-1]
+    index = args[0]
+    # Verify scalar compression was configured
+    assert hasattr(index.vector_search, 'compressions')
+    assert len(index.vector_search.compressions) > 0
+    assert "ScalarQuantizationCompression" in str(type(index.vector_search.compressions[0]))
+    
+    # Test with binary compression
+    instance = AzureAISearch(
+        service_name="test-service",
+        collection_name="binary-index",
+        api_key="test-api-key",
+        embedding_model_dims=768,
+        compression_type="binary"
+    )
+    assert instance.compression_type == "binary"
+    
+    # Capture the index creation call
+    args, _ = mock_index_client.create_or_update_index.call_args_list[-1]
+    index = args[0]
+    # Verify binary compression was configured
+    assert hasattr(index.vector_search, 'compressions')
+    assert len(index.vector_search.compressions) > 0
+    assert "BinaryQuantizationCompression" in str(type(index.vector_search.compressions[0]))
+    
+    # Test with no compression
+    instance = AzureAISearch(
+        service_name="test-service",
+        collection_name="no-compression-index",
+        api_key="test-api-key",
+        embedding_model_dims=768,
+        compression_type=None
+    )
+    assert instance.compression_type == "none"
+    
+    # Capture the index creation call
+    args, _ = mock_index_client.create_or_update_index.call_args_list[-1]
+    index = args[0]
+    # Verify no compression was configured
+    assert hasattr(index.vector_search, 'compressions')
+    assert len(index.vector_search.compressions) == 0
+
+
+def test_initialization_with_float_precision(mock_clients):
+    """Test initialization with different float precision settings."""
+    mock_search_client, mock_index_client, _ = mock_clients
+    
+    # Test with half precision (float16)
+    instance = AzureAISearch(
+        service_name="test-service",
+        collection_name="float16-index",
+        api_key="test-api-key",
+        embedding_model_dims=768,
+        use_float16=True
+    )
+    assert instance.use_float16 is True
+    
+    # Capture the index creation call
+    args, _ = mock_index_client.create_or_update_index.call_args_list[-1]
+    index = args[0]
+    # Find the vector field and check its type
+    vector_field = next((f for f in index.fields if f.name == "vector"), None)
+    assert vector_field is not None
+    assert "Edm.Half" in vector_field.type
+    
+    # Test with full precision (float32)
+    instance = AzureAISearch(
+        service_name="test-service",
+        collection_name="float32-index",
+        api_key="test-api-key",
+        embedding_model_dims=768,
+        use_float16=False
+    )
+    assert instance.use_float16 is False
+    
+    # Capture the index creation call
+    args, _ = mock_index_client.create_or_update_index.call_args_list[-1]
+    index = args[0]
+    # Find the vector field and check its type
+    vector_field = next((f for f in index.fields if f.name == "vector"), None)
+    assert vector_field is not None
+    assert "Edm.Single" in vector_field.type
+
+
+# --- Tests for create_col method ---
+
+def test_create_col(azure_ai_search_instance):
+    """Test the create_col method creates an index with the correct configuration."""
+    instance, _, mock_index_client = azure_ai_search_instance
+    
+    # create_col is called during initialization, so we check the call that was already made
+    mock_index_client.create_or_update_index.assert_called_once()
+    
+    # Verify the index configuration
+    args, _ = mock_index_client.create_or_update_index.call_args
+    index = args[0]
+    
+    # Check basic properties
+    assert index.name == "test-index"
+    assert len(index.fields) == 6  # id, user_id, run_id, agent_id, vector, payload
+    
+    # Check that required fields are present
+    field_names = [f.name for f in index.fields]
+    assert "id" in field_names
+    assert "vector" in field_names
+    assert "payload" in field_names
+    assert "user_id" in field_names
+    assert "run_id" in field_names
+    assert "agent_id" in field_names
+    
+    # Check that id is the key field
+    id_field = next(f for f in index.fields if f.name == "id")
+    assert id_field.key is True
+    
+    # Check vector search configuration
+    assert index.vector_search is not None
+    assert len(index.vector_search.profiles) == 1
+    assert index.vector_search.profiles[0].name == "my-vector-config"
+    assert index.vector_search.profiles[0].algorithm_configuration_name == "my-algorithms-config"
+    
+    # Check algorithms
+    assert len(index.vector_search.algorithms) == 1
+    assert index.vector_search.algorithms[0].name == "my-algorithms-config"
+    assert "HnswAlgorithmConfiguration" in str(type(index.vector_search.algorithms[0]))
+    
+    # With binary compression and float16, we should have compression configuration
+    assert len(index.vector_search.compressions) == 1
+    assert index.vector_search.compressions[0].compression_name == "myCompression"
+    assert "BinaryQuantizationCompression" in str(type(index.vector_search.compressions[0]))
+
+
+def test_create_col_scalar_compression(mock_clients):
+    """Test creating a collection with scalar compression."""
+    mock_search_client, mock_index_client, _ = mock_clients
+    
+    instance = AzureAISearch(
+        service_name="test-service",
+        collection_name="scalar-index",
+        api_key="test-api-key",
+        embedding_model_dims=768,
+        compression_type="scalar"
+    )
+    
+    # Verify the index configuration
+    args, _ = mock_index_client.create_or_update_index.call_args
+    index = args[0]
+    
+    # Check compression configuration
+    assert len(index.vector_search.compressions) == 1
+    assert index.vector_search.compressions[0].compression_name == "myCompression"
+    assert "ScalarQuantizationCompression" in str(type(index.vector_search.compressions[0]))
+    
+    # Check profile references compression
+    assert index.vector_search.profiles[0].compression_name == "myCompression"
+
+
+def test_create_col_no_compression(mock_clients):
+    """Test creating a collection with no compression."""
+    mock_search_client, mock_index_client, _ = mock_clients
+    
+    instance = AzureAISearch(
+        service_name="test-service",
+        collection_name="no-compression-index",
+        api_key="test-api-key",
+        embedding_model_dims=768,
+        compression_type=None
+    )
+    
+    # Verify the index configuration
+    args, _ = mock_index_client.create_or_update_index.call_args
+    index = args[0]
+    
+    # Check compression configuration - should be empty
+    assert len(index.vector_search.compressions) == 0
+    
+    # Check profile doesn't reference compression
+    assert index.vector_search.profiles[0].compression_name is None
+
+
+# --- Tests for insert method ---
+
+def test_insert_single(azure_ai_search_instance):
+    """Test inserting a single vector."""
+    instance, mock_search_client, _ = azure_ai_search_instance
+    vectors = [[0.1, 0.2, 0.3]]
+    payloads = [{"user_id": "user1", "run_id": "run1", "agent_id": "agent1"}]
+    ids = ["doc1"]
+
+    instance.insert(vectors, payloads, ids)
+
+    # Verify upload_documents was called correctly
+    mock_search_client.upload_documents.assert_called_once()
+    args, _ = mock_search_client.upload_documents.call_args
+    documents = args[0]
+    
+    # Verify document structure
+    assert len(documents) == 1
+    assert documents[0]["id"] == "doc1"
+    assert documents[0]["vector"] == [0.1, 0.2, 0.3]
+    assert documents[0]["payload"] == json.dumps(payloads[0])
+    assert documents[0]["user_id"] == "user1"
+    assert documents[0]["run_id"] == "run1"
+    assert documents[0]["agent_id"] == "agent1"
+
+
+def test_insert_multiple(azure_ai_search_instance):
+    """Test inserting multiple vectors in one call."""
+    instance, mock_search_client, _ = azure_ai_search_instance
+    
+    # Create multiple vectors
+    num_docs = 3
+    vectors = [[float(i)/10, float(i+1)/10, float(i+2)/10] for i in range(num_docs)]
+    payloads = [{"user_id": f"user{i}", "content": f"Test content {i}"} for i in range(num_docs)]
+    ids = [f"doc{i}" for i in range(num_docs)]
+    
+    # Configure mock to return success for all documents
+    mock_search_client.upload_documents.return_value = [
+        {"status": True, "id": id_val} for id_val in ids
+    ]
+    
+    # Insert the documents
+    instance.insert(vectors, payloads, ids)
+    
+    # Verify upload_documents was called with correct documents
+    mock_search_client.upload_documents.assert_called_once()
+    args, _ = mock_search_client.upload_documents.call_args
+    documents = args[0]
+    
+    # Verify all documents were included
+    assert len(documents) == num_docs
+    
+    # Check first document
+    assert documents[0]["id"] == "doc0"
+    assert documents[0]["vector"] == [0.0, 0.1, 0.2]
+    assert documents[0]["payload"] == json.dumps(payloads[0])
+    assert documents[0]["user_id"] == "user0"
+    
+    # Check last document
+    assert documents[2]["id"] == "doc2"
+    assert documents[2]["vector"] == [0.2, 0.3, 0.4]
+    assert documents[2]["payload"] == json.dumps(payloads[2])
+    assert documents[2]["user_id"] == "user2"
+
+
+def test_insert_with_error(azure_ai_search_instance):
+    """Test insert when Azure returns an error for one or more documents."""
+    instance, mock_search_client, _ = azure_ai_search_instance
+    
+    # Configure mock to return an error for one document
+    mock_search_client.upload_documents.return_value = [
+        {"status": False, "id": "doc1", "errorMessage": "Azure error"}
+    ]
+    
+    vectors = [[0.1, 0.2, 0.3]]
+    payloads = [{"user_id": "user1"}]
+    ids = ["doc1"]
+    
+    # Insert should raise an exception
+    with pytest.raises(Exception) as exc_info:
+        instance.insert(vectors, payloads, ids)
+    
+    assert "Insert failed for document doc1" in str(exc_info.value)
+    
+    # Configure mock to return mixed success/failure for multiple documents
+    mock_search_client.upload_documents.return_value = [
+        {"status": True, "id": "doc1"},
+        {"status": False, "id": "doc2", "errorMessage": "Azure error"}
+    ]
+    
+    vectors = [[0.1, 0.2, 0.3], [0.4, 0.5, 0.6]]
+    payloads = [{"user_id": "user1"}, {"user_id": "user2"}]
+    ids = ["doc1", "doc2"]
+    
+    # Insert should raise an exception
+    with pytest.raises(Exception) as exc_info:
+        instance.insert(vectors, payloads, ids)
+    
+    assert "Insert failed for document doc2" in str(exc_info.value)
+
+
+def test_insert_with_missing_payload_fields(azure_ai_search_instance):
+    """Test inserting with payloads missing some of the expected fields."""
+    instance, mock_search_client, _ = azure_ai_search_instance
+    vectors = [[0.1, 0.2, 0.3]]
+    payloads = [{"content": "Some content without user_id, run_id, or agent_id"}]
+    ids = ["doc1"]
+
+    instance.insert(vectors, payloads, ids)
+
+    # Verify upload_documents was called correctly
+    mock_search_client.upload_documents.assert_called_once()
+    args, _ = mock_search_client.upload_documents.call_args
+    documents = args[0]
+    
+    # Verify document has payload but not the extra fields
+    assert len(documents) == 1
+    assert documents[0]["id"] == "doc1"
+    assert documents[0]["vector"] == [0.1, 0.2, 0.3]
+    assert documents[0]["payload"] == json.dumps(payloads[0])
+    assert "user_id" not in documents[0]
+    assert "run_id" not in documents[0]
+    assert "agent_id" not in documents[0]
+
+
+def test_insert_with_http_error(azure_ai_search_instance):
+    """Test insert when Azure client throws an HTTP error."""
+    instance, mock_search_client, _ = azure_ai_search_instance
+    
+    # Configure mock to raise an HttpResponseError
+    mock_search_client.upload_documents.side_effect = HttpResponseError("Azure service error")
+    
+    vectors = [[0.1, 0.2, 0.3]]
+    payloads = [{"user_id": "user1"}]
+    ids = ["doc1"]
+    
+    # Insert should propagate the HTTP error
+    with pytest.raises(HttpResponseError) as exc_info:
+        instance.insert(vectors, payloads, ids)
+    
+    assert "Azure service error" in str(exc_info.value)
+
+
+# --- Tests for search method ---
+
+def test_search_basic(azure_ai_search_instance):
+    """Test basic vector search without filters."""
+    instance, mock_search_client, _ = azure_ai_search_instance
+    
+    # Configure mock to return search results
+    mock_search_client.search.return_value = [
+        {
+            "id": "doc1",
+            "@search.score": 0.95,
+            "payload": json.dumps({"content": "Test content"})
+        }
+    ]
+    
+    # Search with a vector
+    query_vector = [0.1, 0.2, 0.3]
+    results = instance.search(query_vector, limit=5)
+    
+    # Verify search was called correctly
+    mock_search_client.search.assert_called_once()
+    _, kwargs = mock_search_client.search.call_args
+    
+    # Check parameters
+    assert len(kwargs["vector_queries"]) == 1
+    assert kwargs["vector_queries"][0].vector == query_vector
+    assert kwargs["vector_queries"][0].k_nearest_neighbors == 5
+    assert kwargs["vector_queries"][0].fields == "vector"
+    assert kwargs["filter"] is None  # No filters
+    assert kwargs["top"] == 5
+    assert kwargs["vector_filter_mode"] == "preFilter"  # Default mode
+    
+    # Check results
+    assert len(results) == 1
+    assert results[0].id == "doc1"
+    assert results[0].score == 0.95
+    assert results[0].payload == {"content": "Test content"}

tests/vector_stores/test_elasticsearch.py

@@ -1,6 +1,6 @@
 import os
 import unittest
-from unittest.mock import MagicMock, patch
+from unittest.mock import MagicMock, Mock, patch
 
 import dotenv
 
@@ -220,6 +220,23 @@ class TestElasticsearchDB(unittest.TestCase):
         self.assertEqual(results[0].score, 0.8)
         self.assertEqual(results[0].payload, {"key1": "value1"})
 
+    def test_custom_search_query(self):
+        # Mock custom search query
+        self.es_db.custom_search_query = Mock()
+        self.es_db.custom_search_query.return_value = {"custom_key": "custom_value"}
+
+        # Perform search
+        query_vector = [0.1] * 1536
+        limit = 5
+        filters = {"key1": "value1"}
+        self.es_db.search(query=query_vector, limit=limit, filters=filters)
+
+        # Verify custom search query function was called
+        self.es_db.custom_search_query.assert_called_once_with(query_vector, limit, filters)
+
+        # Verify custom search query was used
+        self.client_mock.search.assert_called_once_with(index=self.es_db.collection_name, body={"custom_key": "custom_value"})
+
     def test_get(self):
         # Mock get response with correct structure
         mock_response = {

tests/vector_stores/test_weaviate.py

@@ -0,0 +1,220 @@
+import os
+import uuid
+import httpx
+import unittest
+from unittest.mock import MagicMock, patch
+
+import dotenv
+import weaviate
+from weaviate.classes.query import MetadataQuery, Filter
+from weaviate.exceptions import UnexpectedStatusCodeException
+
+from mem0.vector_stores.weaviate import Weaviate, OutputData
+
+
+class TestWeaviateDB(unittest.TestCase):
+    @classmethod
+    def setUpClass(cls):
+        dotenv.load_dotenv()
+
+        cls.original_env = {
+            'WEAVIATE_CLUSTER_URL': os.getenv('WEAVIATE_CLUSTER_URL', 'http://localhost:8080'),
+            'WEAVIATE_API_KEY': os.getenv('WEAVIATE_API_KEY', 'test_api_key'),
+        }
+
+        os.environ['WEAVIATE_CLUSTER_URL'] = 'http://localhost:8080'
+        os.environ['WEAVIATE_API_KEY'] = 'test_api_key'
+
+    def setUp(self):
+        self.client_mock = MagicMock(spec=weaviate.WeaviateClient)
+        self.client_mock.collections = MagicMock()
+        self.client_mock.collections.exists.return_value = False
+        self.client_mock.collections.create.return_value = None
+        self.client_mock.collections.delete.return_value = None
+
+        patcher = patch('mem0.vector_stores.weaviate.weaviate.connect_to_local', return_value=self.client_mock)
+        self.mock_weaviate = patcher.start()
+        self.addCleanup(patcher.stop)
+
+        self.weaviate_db = Weaviate(
+            collection_name="test_collection",
+            embedding_model_dims=1536,
+            cluster_url=os.getenv('WEAVIATE_CLUSTER_URL'),
+            auth_client_secret=os.getenv('WEAVIATE_API_KEY'),
+            additional_headers={"X-OpenAI-Api-Key": "test_key"},
+        )
+
+        self.client_mock.reset_mock()
+
+    @classmethod
+    def tearDownClass(cls):
+        for key, value in cls.original_env.items():
+            if value is not None:
+                os.environ[key] = value
+            else:
+                os.environ.pop(key, None)
+
+    def tearDown(self):
+        self.client_mock.reset_mock()
+
+    def test_create_col(self):
+        self.client_mock.collections.exists.return_value = False
+        self.weaviate_db.create_col(vector_size=1536)
+
+
+        self.client_mock.collections.create.assert_called_once()
+
+
+        self.client_mock.reset_mock()
+
+        self.client_mock.collections.exists.return_value = True
+        self.weaviate_db.create_col(vector_size=1536)
+
+        self.client_mock.collections.create.assert_not_called()
+
+    def test_insert(self):
+        self.client_mock.batch = MagicMock()
+
+        self.client_mock.batch.fixed_size.return_value.__enter__.return_value = MagicMock()
+        
+        self.client_mock.collections.get.return_value.data.insert_many.return_value = {
+            "results": [{"id": "id1"}, {"id": "id2"}]
+        }
+        
+        vectors = [[0.1] * 1536, [0.2] * 1536]
+        payloads = [{"key1": "value1"}, {"key2": "value2"}]
+        ids = [str(uuid.uuid4()), str(uuid.uuid4())]
+        
+        results = self.weaviate_db.insert(vectors=vectors, payloads=payloads, ids=ids)
+
+    def test_get(self):
+        valid_uuid = str(uuid.uuid4())
+
+        mock_response = MagicMock()
+        mock_response.properties = {
+            "hash": "abc123",
+            "created_at": "2025-03-08T12:00:00Z",
+            "updated_at": "2025-03-08T13:00:00Z",
+            "user_id": "user_123",
+            "agent_id": "agent_456",
+            "run_id": "run_789",
+            "data": {"key": "value"},
+            "category": "test",
+        }
+        mock_response.uuid = valid_uuid
+
+        self.client_mock.collections.get.return_value.query.fetch_object_by_id.return_value = mock_response
+
+        result = self.weaviate_db.get(vector_id=valid_uuid)
+
+        assert result.id == valid_uuid
+        
+        expected_payload = mock_response.properties.copy()
+        expected_payload["id"] = valid_uuid
+
+        assert result.payload == expected_payload
+
+
+    def test_get_not_found(self):
+        mock_response = httpx.Response(status_code=404, json={"error": "Not found"})
+
+        self.client_mock.collections.get.return_value.data.get_by_id.side_effect = UnexpectedStatusCodeException(
+            "Not found", mock_response
+        )
+
+
+    def test_search(self):
+        mock_objects = [
+            {
+                "uuid": "id1",
+                "properties": {"key1": "value1"},
+                "metadata": {"distance": 0.2}
+            }
+        ]
+        
+        mock_response = MagicMock()
+        mock_response.objects = []
+        
+        for obj in mock_objects:
+            mock_obj = MagicMock()
+            mock_obj.uuid = obj["uuid"]
+            mock_obj.properties = obj["properties"]
+            mock_obj.metadata = MagicMock()
+            mock_obj.metadata.distance = obj["metadata"]["distance"]
+            mock_response.objects.append(mock_obj)
+        
+        mock_hybrid = MagicMock()
+        self.client_mock.collections.get.return_value.query.hybrid = mock_hybrid
+        mock_hybrid.return_value = mock_response
+        
+        query_vector = [0.1] * 1536
+        results = self.weaviate_db.search(query=query_vector, limit=5)
+        
+        mock_hybrid.assert_called_once()
+        
+        self.assertEqual(len(results), 1)
+        self.assertEqual(results[0].id, "id1")
+        self.assertEqual(results[0].score, 0.8)
+
+    def test_delete(self):
+        self.weaviate_db.delete(vector_id="id1")
+
+        self.client_mock.collections.get.return_value.data.delete_by_id.assert_called_once_with("id1")
+
+    def test_list(self):
+        mock_objects = []
+        
+        mock_obj1 = MagicMock()
+        mock_obj1.uuid = "id1"
+        mock_obj1.properties = {"key1": "value1"}
+        mock_objects.append(mock_obj1)
+        
+        mock_obj2 = MagicMock()
+        mock_obj2.uuid = "id2"
+        mock_obj2.properties = {"key2": "value2"}
+        mock_objects.append(mock_obj2)
+        
+        mock_response = MagicMock()
+        mock_response.objects = mock_objects
+        
+        mock_fetch = MagicMock()
+        self.client_mock.collections.get.return_value.query.fetch_objects = mock_fetch
+        mock_fetch.return_value = mock_response
+        
+        results = self.weaviate_db.list(limit=10)
+        
+        mock_fetch.assert_called_once()
+        
+        # Verify results
+        self.assertEqual(len(results), 1)
+        self.assertEqual(len(results[0]), 2)
+        self.assertEqual(results[0][0].id, "id1")
+        self.assertEqual(results[0][0].payload["key1"], "value1")
+        self.assertEqual(results[0][1].id, "id2")
+        self.assertEqual(results[0][1].payload["key2"], "value2")
+
+
+    def test_list_cols(self):
+        mock_collection1 = MagicMock()
+        mock_collection1.name = "collection1"
+
+        mock_collection2 = MagicMock()
+        mock_collection2.name = "collection2"
+        self.client_mock.collections.list_all.return_value = [mock_collection1, mock_collection2]
+
+        result = self.weaviate_db.list_cols()
+        expected = {"collections": [{"name": "collection1"}, {"name": "collection2"}]}
+
+        assert result == expected
+
+        self.client_mock.collections.list_all.assert_called_once()
+
+
+    def test_delete_col(self):
+        self.weaviate_db.delete_col()
+
+        self.client_mock.collections.delete.assert_called_once_with("test_collection")
+
+
+if __name__ == '__main__':
+    unittest.main()