One place for your notes, decisions, and workflows. For you, your AI tools, and your team.
Connected in a living knowledge graph. Plain markdown you control.
## What it actually is
Built-in AI memory is small, vendor-owned, and usually invisible. Basic Memory is large, structured, and inspectable: a knowledge graph written in plain Markdown. Open it in any editor, hand it to any AI tool, and keep it, because the files are yours.
## Why it's different
The long-term part of AI memory belongs in plain Markdown files you own. They stay readable, portable, and yours. Keeping the files is the easy half. The hard half is making them instantly searchable without locking them away. That's the part we built.
Your memory is embeddings in a store you can't open. Great for the machine, opaque to you. And gone if you leave.
Readable and yours, but the AI just greps or reloads whole files. No graph, no semantic search, doesn't scale.
Plain Markdown you own, with a full-text, semantic, and knowledge-graph index on top. The files stay plain; the search gets smart.
## Made to work together
Built-in memory is good at short-term. What you said yesterday, your preferences, the thread of a chat. Keep using it. We're the layer underneath: a deep, owned knowledge base for the things that matter longer than a session. Codebase decisions, research, projects, a real second brain. The short-term layer remembers; the long-term layer compounds.
## What you can do with it
A folder of Markdown is a great start. We've all built one. But the moment an agent has to actually use it, the questions get hard. How does it search? Follow the graph? Know what changed? Stay consistent across people and tools? That's what the index is for, and that's the part we built.
Find by meaning, not just keywords. Vector and hybrid search across everything you've written.
Filter by frontmatter. Status, type, tags, priority, or any custom field. Query your notes like records.
Define and validate note types. Task, Person, Client, Ticket. So structure holds as the graph grows.
Typed relations and observations connect notes into a graph your AI can actually traverse.
Drop-in setups for Claude, Codex, OpenClaw, and Hermes. Search-before-answer recall and auto-capture.
Reusable bundles of schemas, templates, and agent instructions you can deploy into any workspace.
## Start where you are
Run Basic Memory locally, sync a private cloud knowledge base, or scale it across a team and its agents. Same markdown, same graph, same ownership model. You just choose how far it reaches.
Open source and offline. Your knowledge graph lives on your own machine. 3K+ stars and 57K downloads a month.
For buildersPrivate sync for a knowledge base that follows you across devices and AI tools. Starts with a 7-day trial; seats get cheaper as a team grows.
See pricingCredit-based usage for agent fleets and automation, with enterprise, custom-deployment, and on-prem options.
Agent memory playbook## The asset is the knowledge base
Basic Memory keeps the work that matters in a knowledge base you own. Plain Markdown files you can read, edit, and back up like any folder. It outlasts any single chat, app, or subscription.
Keep decisions, research, notes, and project history in files you can open, search, and edit by hand.
Give Claude, Codex, Cursor, ChatGPT, and any MCP-compatible tool the same structured context to work from.
Build one shared body of knowledge instead of scattering context across chats, docs, and people.
## No lock-in
One knowledge graph, read and written through the Model Context Protocol. No lock-in, no proprietary store. Just Markdown files and an open standard the whole ecosystem speaks. Switch between Claude, ChatGPT, and Cursor whenever you want, and your memory comes with you.
## Built to share
Start solo and local. When you're ready to share, the same knowledge graph scales. To a team, to the clients an MSP manages, to whoever needs context. Same Markdown. Same ownership. Same files.
A shared workspace where people and AI agents read and write the same knowledge graph. With roles, live collaborative editing, and audit logs.
Explore TeamsDeploy a private, isolated knowledge base for every client. And let Claude run the back office against it. One platform, separate tenants, consolidated billing.
Partner programStart with a 7-day trial. Talk to us when Basic Memory becomes infrastructure.
7-Day Trial
Add or remove seats anytime.
Monthly pricing is calculated automatically at checkout.
Agent Infrastructure
For agent fleets, custom deployment, enterprise, and on-prem needs.
Good fit when Basic Memory becomes production memory infrastructure.
Deploying Basic Memory for clients? Learn about the partner program →
Prefer to self-host? The full app is free and open source (AGPL-3.0). Complete control, complete privacy.
## Get started
Free and open source to run locally. Two minutes to connect your first AI tool. Your files, your graph, forever.
Knowledge tools built to last
© 2026 Basic Memory - Basic Machines
Claude
Codex
Cursor
Obsidian
VS Code
MCP
Markdown