Mem0 reads a conversation, decides what is worth keeping, stores it as searchable facts, and finds them again next time the same person turns up. It is two method calls with a model and an embedder behind them, and it is the piece that stops an assistant asking the same question twice.
Reach for it when
- Your assistant asks users for the same details every session
- You need per-user facts that outlive one conversation
- You want memory you can inspect, update and delete
Look elsewhere when
- You only need the last few turns of one chat. That is a context window
- You want to search documents rather than remember people. That is RAG
What the lessons will cover
- 01Why a dictionary of notes is not memory
- 02A stand-in model and embedder
- 03add, search and filters
- 04user_id, agent_id and run_id
- 05Metadata and filtering
- 06Updates, deletion and history
- 07Custom instructions
- 08Where memories live
- 09An assistant that remembers, tested
Others in memory
Common questions
Is Mem0 free to use?
Mem0 is open source and free to run yourself. You still pay whichever model provider you point it at, and this tutorial is free with no signup.
Do I need to know Python to use Mem0?
Basic Python is enough. Functions, dictionaries and imports cover most of what Mem0 asks of you.
When should I not use Mem0?
You only need the last few turns of one chat. That is a context window. You want to search documents rather than remember people. That is RAG.
Lessons for Mem0 are being written. Meanwhile the Agent Governance Toolkit tutorial covers the same ground: state, tools, loops, memory and human approval. Most of it carries straight over.
Start the Agent Governance Toolkit tutorial →