Mem0 Open Source
Self-hostable memory engine for AI applications, available as a library or a server with a dashboard and per-user API keys.
About this listing
Mem0 Open Source gives AI application developers a memory engine they can run on their own infrastructure. It can be embedded as a library in an application or deployed as a self-hosted server. The server option includes a dashboard, per-user API keys, and a request audit log.
Developers can configure the language model, embedding model, vector store, and reranker rather than treating memory as a fixed hosted dependency. Python and Node.js starting guides cover the library route, while a Docker-based server setup supports teams that want a shared service.
Free refers to the Apache-2.0-licensed source code. It does not include free model inference, embeddings, storage, or infrastructure. The documented defaults use OpenAI models, so review provider charges and configuration before deployment. This listing is for the self-hosted open-source engine, not the separately managed Mem0 Platform.