Verified

Mem0 MCP Server

Persistent memory for LLMs using the Mem0 API for cross-session data recall.

Data last scanned yesterday · Reviewed last month

13 2 dangerous tools16 tools★ 99TypeScriptMIT

Overview

This server provides tools to interface with Mem0, enabling AI agents to store, update, search, and list persistent memories. It uses the mem0ai Node.js SDK to manage context across sessions, allowing for advanced retrieval through semantic and BM25 hybrid search functionality. The server supports granular scoping by user, agent, and application identifiers. Developers can perform asynchronous memory additions with configurable polling or retrieve specific memory records by ID to refine the information context available to their models.

Our verdict

With a security score of 43/100, this server is classified as high risk. The evaluation process has not yet verified the authentication methods or the read-only mode, and the project is maintained by community members rather than the official Mem0 vendor. Because of these unreviewed security dimensions, it should be used cautiously in environments that handle sensitive user data.

Supports semantic and BM25 hybrid memory searching
Small footprint with only two direct dependencies
Provides paginated memory listing and record retrieval
Active development with commits as recent as three days ago

Authentication methods have not been reviewed
Read-only mode support remains unverified
Not maintained by the official Mem0 organization

Setup

Claude Desktop

{
  "mcpServers": {
    "mem0-mcp": {
      "args": [
        "-y",
        "@pinkpixel/mem0-mcp"
      ],
      "command": "npx"
    }
  }
}

Paste into ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows), then restart Claude Desktop. Some servers require API keys or tokens as environment variables — check the project's README and add an "env" object if needed.

No read-only mode — running this grants full read/write access.
No built-in authentication — anyone with access to this config can invoke every tool it exposes.

Tools

Tool	Description	Risk
add_memory	Stores a memory from text content or structured message arrays.	medium
search_memories	Searches memories using semantic and BM25 hybrid filters.	low
search_memory	Backward-compatible alias for search_memories.	low
list_memories	Paginated listing of memory records scoped by identifiers.	low
get_memory	Retrieves a single memory record by its ID.	low
update_memory	Modifies the text or metadata of an existing memory.	medium
delete_memory	Deletes a specific memory record by ID.	high
get_memory_history	Retrieves the audit trail of memory revisions (cloud only).	low
get_memory_capabilities	Exposes the feature matrix and support flags of the active backend storage mode.	low
batch_update_memories	Performs bulk updates of text contents for multiple memories (cloud only).	medium

Compatibility

Client	Local	Docker	Remote	Read-only
ChatGPT
Claude Desktop
Cursor
VS Code
Windsurf

Frequently asked questions

›What are the storage modes available for this server?

This server supports Cloud Storage using a Mem0 API key, Supabase Storage for self-hosting with your own database and OpenAI credentials, and Local Storage for development and testing.

›What is the behavior of the add_memory tool regarding background tasks?

By default, the tool polls the background queue until the operation is completed. You can set the waitForCompletion parameter to false if you prefer to receive the eventId immediately.

›Which features are exclusive to the Cloud Storage mode?

Cloud-only features include memory history tracking, batch updates, memory rating/feedback, and memory export job management.

›Is data persistence guaranteed in Local Storage mode?

No, Local Storage mode uses an in-memory vector database and data is lost when the server restarts unless specifically configured otherwise.

›What prerequisites are required for Supabase Storage mode?

You must provide SUPABASE_URL, SUPABASE_KEY, and an OPENAI_API_KEY for embedding generation, as well as perform the necessary initial database SQL migrations.

›Are there any safety requirements for the batch_delete_memories tool?

Yes, to prevent accidental deletion, the confirm parameter must be explicitly set to true to execute a batch deletion.