Overview
Supermemory is a memory engine and end-user application built to store, organize, and let you chat with the information that matters. It is positioned as an AI-era “second brain”: you can add memories from URLs, PDFs, notes, and files, then query or chat with that content using natural language. The project provides both consumer-facing apps (web app, browser extension, Raycast extension) and developer-facing capabilities (Developer Console, API, documentation, and self-hosting guides).
Key features
- Add memories from multiple content types: URLs, PDFs, plain text, notes, and connected services (Notion, Google Drive, OneDrive).
- Conversational retrieval: open a chat over your stored memories to ask questions and retrieve context-aware answers.
- Supermemory MCP integration: connect Supermemory to a variety of AI tools and LLMs via MCP to enable downstream model access to stored memories.
- Browser extension (Chrome/Edge): save content directly from webpages, integrate with Chat UI flows, and import from social platforms.
- Raycast extension: quickly add and search memories using keyboard shortcuts.
- Developer API & Console: API access and developer console for building on top of the memory engine.
- Self-hosting & enterprise deployment: documentation and guides for deploying Supermemory on your own infrastructure.
Integrations & extensions
Supermemory emphasizes interoperability. It offers official integrations and connectors (Notion, Google Drive, OneDrive) and supports connecting to multiple AI tools through MCP. Extensions include a browser extension for one-click saving from web pages and a Raycast plugin for fast keyboard-driven access. The repository links to the web app, Chrome extension, Raycast page, and Discord community.
Intended users & use cases
- Individual users who want a personal knowledge base and a searchable conversational memory.
- Product teams building RAG-enabled applications that need a memory store for context retrieval.
- Enterprises requiring a deployable memory layer with self-hosting and API access for compliance and scale.
- Developers building AI assistants who need a fast, scalable memory API to persist and query user data.
Developer & deployment notes
The project includes a Developer Console and documentation for API usage. There is also a self-hosting guide tailored for enterprise deployments. The README highlights contribution guidelines and invites community involvement (issues, good first issues, help wanted labels).
How to get started
- Visit the web app (app.supermemory.ai) and sign up.
- Add memories via paste/upload or by connecting external services.
- Open Chat to interact with your saved memories or connect Supermemory to an LLM via MCP.
- Developers can use the Developer Console and API for programmatic access or follow self-hosting docs for enterprise setups.
Notes
Supermemory aims to be both a consumer product (web and extensions) and a developer platform (API, self-hosting). Its architecture and integration points make it suitable for RAG workflows and for augmenting LLMs with persistent, queryable context.
