All open source
Open SourceRAG40k

mem0

Memory layer for agents; auto-extracts facts into vector and graph.

Mem0 is an open-source memory layer for AI agents and assistants that extracts, stores, and retrieves long-term memories across conversations and sessions. It identifies salient facts from interactions, persists them in vector and graph stores, and surfaces relevant memories to personalize future responses, exposed through a Python/JS SDK and a hosted platform. It is notable for focusing specifically on persistent, personalized agent memory rather than general document retrieval.

Repository

mem0ai/mem0

Language

Python

What you'd build with it

  • Giving a chatbot or assistant persistent memory of user preferences and past conversations
  • Personalizing agent responses by retrieving relevant stored facts at inference time
  • Maintaining cross-session context for multi-turn agents without manually managing memory state

Tags

memoryagentspersonalization