supermemoryai/supermemory is a memory and context engine for AI, and its latest release is server-v0.0.9. The repository description says it provides both an engine and an app, and that it can run completely locally. It extracts facts from conversations and manages user profiles and knowledge changes, so it can supply that context whenever it is needed.
The core point is that a single Memory API bundles RAG, hybrid search, connectors, and file handling together. RAG is the approach of combining evidence retrieved from outside with the model's answer. Attaching a plugin or MCP server gives compatible AI assistants persistent memory, and local execution and self-hosting are supported as well. The license is MIT, the language is TypeScript, and the repository has 31342 stars.
That said, the release notes for server-v0.0.9 are not in this record, so what this version fixed cannot be confirmed. The fast processing and scalability emphasized in the repository description are the project's own claims, and no comparison performance figures are in this record either. Among the collected evaluation scores, only the Momentum entry is 5.5, while the freshness, technical, and quality entries are empty. Considering that the repository was created on February 27, 2024, it remains a candidate worth trying for anyone who wants to attach a memory layer directly. Whether this release brings substantive improvements can only be judged by checking the release notes directly.