Supermemory is a memory and context layer for AI applications. It extracts facts from conversations, maintains a profile for each user, handles knowledge updates and contradictions, expires outdated information, and returns relevant context at query time. Retrieval-augmented generation (RAG) over documents, data connectors, and file processing are part of the same system and share one memory structure and ontology.
There are three ways to use it. If you use AI tools, the consumer app, browser extension, plugins, and hosted MCP server give assistants such as Claude Desktop, Cursor, Windsurf, VS Code, Claude Code, OpenCode, OpenClaw, and Hermes persistent memory through three tools: memory (save or forget), recall (search plus profile summary), and context (inject the full profile at session start). If you build AI products, the API and the npm and PyPI SDKs add memory, RAG, profiles, and connectors to your own agents. If you want to run it yourself, a single binary (supermemory-server) serves the same API on localhost:6767 with local embeddings and any OpenAI-compatible model, including fully offline through Ollama.
Memory is scoped with container tags, so work and personal context, or one client and another, stay separate. A profile call returns static long-term facts and dynamic recent context in one response; search runs in hybrid mode (documents plus memories) or memories-only mode.
Features
- Memory extraction: facts pulled from conversations, with temporal changes, contradictions, and automatic forgetting handled
- User profiles: auto-maintained static and dynamic context per container tag, returned in one call
- Hybrid search: knowledge-base documents and personalised memories answered from a single query
- Connectors: Google Drive, Gmail, Notion, OneDrive, GitHub, and a web crawler, synced through real-time webhooks
- Multi-modal extractors: PDFs, images with OCR, videos with transcription, and code with AST-aware chunking
- Framework wrappers: Vercel AI SDK, LangChain, LangGraph, OpenAI Agents SDK, Mastra, Agno, the Claude Memory Tool, and n8n
- Open-source plugins: Claude Code, OpenCode, OpenClaw, and a Hermes memory provider
- Local deployment: one binary with an embedded graph engine, local bge-base-en-v1.5 embeddings, and a first-boot wizard for model setup
Integrates with
Agno
SDK, AgentOS runtime, and web UI for building and operating multi-tenant agent platforms on your own infrastructure
Claude Code
Terminal-based agentic coding tool that reads your codebase and runs tasks through natural-language commands
Hermes
Self-improving AI agent with a learning loop that creates and refines skills, recalls past sessions, and runs across chat platforms
LangChain
Framework for building LLM applications and agents from interoperable components and a large integrations library
Integrated by
Alternatives
Mem0
Memory layer that extracts facts from conversations and retrieves them at inference so agents remember users across sessions
Zep
Examples, framework integrations, and ingestion tooling for the Zep Cloud agent memory platform
Agentmemory
#1 Persistent memory for AI coding agents based on real-world benchmarks
Iai Personal Memory Engine
Independent Autistic Intelligence — a cyber brain for your AI. It never forgets a detail, remembers exactly what you said, and learns how yo