Agent memory system that stores, recalls and reflects on memories so AI agents learn over time instead of only replaying chat history.
- Stars
- 31,327
- Forks
- 2,750
- Open issues
- 34
- Last commit
- 3 days ago
What is it?
Supermemory is a memory and context layer for AI. It extracts facts from conversations, tracks updates and contradictions, forgets expired information, and maintains a user profile with stable facts and recent activity. Developers can call one API to store content, search across both personal memories and uploaded documents in a hybrid query, and fetch a profile for a given user.
For developers, it ships SDKs for TypeScript and Python, wrappers for frameworks such as the Vercel AI SDK, LangChain, LangGraph and Mastra, and connectors that sync Google Drive, Gmail, Notion, OneDrive, GitHub and web pages. File processing covers PDFs, images via OCR, video transcription and code chunking.
For individual users, it provides plugins for Claude Code, Cursor, Codex, OpenCode and other AI tools, plus an MCP server at a hosted URL. A self-hosted local mode runs the same API on localhost with local embeddings and optional Ollama models, and Docker images are available for Linux amd64 and arm64.
Key features
- 01Extracts facts from conversations and handles updates, contradictions and expiry.
- 02Maintains user profiles with static facts and recent dynamic context in one call.
- 03Hybrid search combines RAG over documents with personalized memory in a single query.
- 04Connectors sync Google Drive, Gmail, Notion, OneDrive, GitHub and web pages.
- 05Processes PDFs, images, videos and code into searchable chunks.
- 06Runs locally with the same API, using local embeddings or Ollama.
Use cases
- Adding persistent memory and personalization to a chatbot or agent.
- Giving AI coding assistants like Cursor or Claude Code memory of preferences and projects.
- Building a RAG knowledge base that syncs from Notion, Drive or GitHub.
- Running a private memory service on your own machine or server.
Quick start
- 1
Install the SDK with npm
npm install supermemory - 2
Create a client and call add and profile as shown in the Quickstart
Things to know
- The hosted API needs a Supermemory account and API key.
- Running locally requires an LLM provider key or a local Ollama setup.
- Local mode downloads an embedding model on first boot.
Topics
FAQ
- What is supermemory?
- Memory and context engine for AI apps and assistants, offering a single API for persistent memory, RAG, user profiles and connectors. Supermemory is a memory and context layer for AI. It extracts facts from conversations, tracks updates and contradictions, forgets expired information, and maintains a user profile with stable facts and recent activity. Developers can call one API to store content, search across both personal memories and uploaded documents in a hybrid query, and fetch a profile for a given user.
- Is supermemory free to use? What license does it have?
- supermemory is open source under the MIT license. Check the license terms before using it in commercial products.
- Is supermemory still maintained?
- Actively maintained. The last commit to supermemory was on Oct 18, 2026.
- What can I use supermemory for?
- Adding persistent memory and personalization to a chatbot or agent. Giving AI coding assistants like Cursor or Claude Code memory of preferences and projects. Building a RAG knowledge base that syncs from Notion, Drive or GitHub. Running a private memory service on your own machine or server.
- How do I install supermemory?
- Install the SDK with npm: npm install supermemory Create a client and call add and profile as shown in the Quickstart
- Where can I get supermemory?
- The source code and installation instructions are on GitHub: https://github.com/supermemoryai/supermemory
Summary generated by AI from the project's README. GitHub stats refresh daily.
Related projects
Agent-native memory infrastructure that turns LLM agent conversations and execution into persistent, structured memory for Python and TypeScript apps.
A memory layer for AI agents and apps that stores and retrieves user, session, and agent context across conversations.
Local-first AI memory system that stores conversation history verbatim and retrieves it with semantic search, exposed as an MCP server and CLI.
Self-hosted AI chat and knowledge platform that indexes data from 50+ connected apps and works with most LLM providers.
Open-source AI memory platform that turns documents, code and conversations into a self-hosted knowledge graph for AI agents.