Mem0

About

Mem0 offers a powerful, self-hosted solution for managing AI memory, integrating seamlessly with the Model Context Protocol. It leverages PostgreSQL and pgvector for robust, scalable memory storage, allowing users to maintain full control over their data. The server supports multiple LLM providers, including Gemini, OpenAI, and Anthropic, with flexible fallback chains for both LLM and embedding models, ensuring reliable AI interactions. Designed for zero manual setup, it simplifies deployment, requiring only a database URL and API keys to get started.

Key Features

• Multi-provider LLM support (Gemini, OpenAI, Anthropic, Groq, DeepSeek, Mistral)
• Self-hosted PostgreSQL for data sovereignty (Neon/Supabase supported)
• 1 GitHub stars
• Configurable fallback chains for LLM models and API keys
• Zero-manual setup requiring only DATABASE_URL and API_KEYS
• Graph memory for SQL-based relationship tracking alongside vector memory

Use Cases

• Providing persistent, self-managed memory for AI agents and applications
• Storing user preferences and contextual information for personalized AI interactions
• Implementing complex knowledge graphs and relationships for AI memory retrieval