Mem0 MCP

Mem0 MCP is an MCP server that integrates with Mem0.ai to provide persistent memory capabilities for LLMs and AI agents across sessions. It connects LLM interfaces to Mem0 Cloud, a self-hosted Supabase database, or a local in-memory store. Developers building interactive coding assistants, customer support agents, and personal productivity tools use this server to retain context, preferences, and factual notes over time. The server provides a full suite of memory management tools, allowing agents to store memories from plain text or conversational message arrays, update contents, manage deletions, and query information using semantic and BM25 hybrid search. When using Mem0 Cloud, it further supports features such as memory version history tracking, asynchronous export query jobs, background event status inspection, and user rating evaluations. By decoupling short-term conversational context from long-term memory, the server ensures assistants recall relevant user specifics while avoiding context window overflow.

Category: AI Memory & Context

Tags: context, mem0, persistent-memory, rag, supabase

Visit Mem0 MCP

How to install and configure Mem0 MCP

1. Install the server package globally via npm or run it directly using npx: bash npm install -g @pinkpixel/mem0-mcp Alternatively, install via Smithery for Claude Desktop: bash npx -y @smithery/cli install @pinkpixel-dev/mem0-mcp-server --client claude 2. Open your MCP client configuration file (such as Claude Desktop or Cursor configuration) and add the server definition under mcpServers using your preferred storage backend. For Mem0 Cloud mode: json { "mcpServers": { "mem0-mcp": { "command": "npx", "args": ["-y", "@pinkpixel/mem0-mcp"], "env": { "MEM0_API_KEY": "YOUR_MEM0_API_KEY_HERE", "DEFAULT_USER_ID": "user123" } } } } For Supabase mode, supply SUPABASE_URL, SUPABASE_KEY, and OPENAI_API_KEY in the env object instead. 3. Save the configuration and restart your client to enable the memory tools.

What you can do with Mem0 MCP

• Storing user coding preferences and library choices across distinct developer chat sessions in Claude Desktop or Cursor. - Searching past interaction notes using hybrid semantic and BM25 filters to answer continuity questions without prompt bloat. - Running batch memory updates or deletions to curate outdated project context across long-running autonomous development workflows. - Tracking memory revision history and logging evaluation feedback on individual memory records using Mem0 Cloud features.

Key facts

• https://github.com/pinkpixel-dev/mem0-mcp
• AI Memory & Context, Databases & Data Stores
• context, mem0, persistent-memory, rag, supabase

Part of MCP Servers

How do I install Mem0 MCP?

You can install Mem0 MCP globally using npm install -g @pinkpixel/mem0-mcp, run it directly via npx @pinkpixel/mem0-mcp, or install it automatically for Claude Desktop using the Smithery CLI command. You then configure the executable and required environment variables in your client configuration file.

What storage backends does Mem0 MCP support?

Mem0 MCP supports three storage backends: Mem0 Cloud mode using an official Mem0 API key, Supabase mode for self-hosting with Supabase credentials and an OpenAI API key, and a local development mode running an in-memory vector database with an OpenAI API key.

What tools are available in Mem0 MCP?

The server provides tools such as add_memory, search_memories, list_memories, get_memory, update_memory, and delete_memory. When running in Mem0 Cloud mode, it adds tools for revision history, batch updates, ratings, background event monitoring, and memory exports.

Which MCP clients work with Mem0 MCP?

Mem0 MCP works with any Model Context Protocol compliant host client that supports stdio process execution, including Claude Desktop, Cursor, Cline, and Roo Code.

Is Mem0 MCP open source?

Yes, Mem0 MCP is open source software released under the MIT license, with the source code publicly hosted on GitHub.