Mem0
A memory layer for retaining and retrieving information across agent interactions. It offers a library, a self-hosted server, and a managed platform with distinct feature scopes.
Utility: Knowledge and memory; kind: infrastructure.
Stars: 68k+; Starred in September: 2k+.
Metrics updated: 31 Oct 2026. Profile reviewed: 31 Oct 2026.
Utilities
- Store user- or agent-specific context for later retrieval.
- Add persistent memory to an existing application.
Who it is for
- Developers who need memory infrastructure around their own agents.
Requirements
- Install the Python library, deploy the server, or configure managed-platform access.
- Configure model, embedding, storage, and authentication settings.
Limitations
- Memory extraction and retrieval can retain incorrect or outdated information.
- The README's managed-platform benchmarks include proprietary optimizations absent from the open-source SDK.
Cost and privacy notes
- Model extraction, embeddings, storage, and hosted usage can add separate charges.
- Persistent memories need consent, user isolation, deletion, and retention policies.
Illustrative example (untested)
Illustrative, not tested: store a user's explicitly approved preference and check what the agent retrieves in a later session.
License
Apache-2.0 for the public repository; managed-platform optimizations and services have separate scope.