supermemory

Supermemory gives your AI assistant a long-term memory that survives across conversations. It extracts facts from chats, keeps a profile of who you are and what you're working on, and runs hybrid retrieval that combines RAG with personalized recall. Plug it into your own agent via the API, or use the consumer app and browser extension to layer memory on top of ChatGPT and Claude.

Self-hosted from the MIT source. You install via `pip install supermemory` or `npm install supermemory`, or wire it into an MCP-compatible client with a one-line install command. It can sync with Google Drive, Gmail, Notion, GitHub, and others, and it ingests PDFs, images, videos, and code as first-class objects.

Solo developers building agents: this is a faster path than rolling your own memory layer. Small teams: the consumer app and developer SDK are both free; production deployments at any scale will likely want the hosted API for managed retrieval and the syncing infrastructure. The consumer app at app.supermemory.ai is free for end-users.

The catch: this is a young, fast-moving project from a research-lab style organization. The abstraction is right but expect breaking changes. The hosted API pricing isn't published; treat any production commitment as a conversation, not a click-to-buy.