Chrome extension. Install Now.
Cairn is a Chrome extension for people working through real university lecture series — Stanford CS, MIT OCW, and the rest. It tracks your progress across devices, and puts a study chat beside the video that knows the lecture you are on — and, where the course has a knowledge base, has already done the reading.
How it works
Track your progress automatically.Cairn watches what you watch on YouTube and tracks it against your enrolled courses. There is no "mark complete" button to click. Sign in with Google and your progress follows you to the next computer, picking up where you left off — at the second you left off. No account needed for tracking if you'd rather keep things local.
Ask about the lecture, right beside it.The study chat in the sidebar knows the course, the lecture, and the second you are paused at. Capture the frame on screen with one click, and "this slide" means this slide.
Answers from the course itself.Courses can carry a knowledge base built from their own transcripts, slides, and handouts. The chat reads it before answering and links its sources, so you can check them yourself.
Study chat
Cairn's study chat sits in the sidebar, next to the video. It already knows which course you are in, which lecture is playing, and the second you are paused at — so you can point at something with "this" and it knows what you mean. On a course with a knowledge base, it can quote the lecture and the slides too.
Bring your own key
The chat is off until you sign in and add your own API key from DeepSeek, OpenRouter, or GLM, so you pay the model provider directly at cost — Cairn takes no cut and resells nothing. Add a Tavily key as well and it can search the web when the course materials run out. Your keys are stored encrypted and never sent back to your browser, and your conversations are kept in your browser rather than on Cairn's servers. The privacy policy explains exactly where each message goes.
Knowledge base
A general chatbot is vague about a specific course. So courses on Cairn can carry a knowledge base, built from that course's own materials.
Each one is a public GitHub repository: every lecture transcript, timestamped; the slide decks and handouts from the course website; and topic pages compiled from them and linked together. When a course has one, the chat reads it before answering, treats it as more reliable than its own memory, and links the file it used so you can read the source yourself. If a lecture assumes background that another course teaches, it can open that course's knowledge base too.
Look at one. The knowledge base for Stanford's CS224N — Natural Language Processing with Deep Learning — is public: chaimantec/cairn-kb-cs224n. All 23 lectures, with timestamped transcripts, the full text of every slide deck, and 86 interlinked topic pages.
You don't have to use Cairn to read it. A knowledge base is ordinary markdown in a public Git repository, so you can browse it on GitHub, clone it, or point Claude Code, Codex, or any other agent at it and work through the course there instead. Nothing in it is locked to the extension — which also means you can check what the assistant is working from, and correct it.
More are being built. Stanford's CS336, CS221, and CS329A and MIT's 6.7960 have knowledge bases too, some still being filled in lecture by lecture. Whoever added a course to the catalog can attach a knowledge base to it, and courses that have one link to it from the catalog.
What's different
Most AI study tools start from a chatbot and bolt the material on. Cairn starts from the lectures — what you have watched, where you are now, what the course actually taught — and keeps the assistant answerable to them.
Alongside the chat, you can pin notes to any second of a lecture to come back to later. Make one public and the next learner sees it at that same moment, under an anonymous nickname you choose. There are no feeds, no followers, and no DMs.
The course catalog grows from the people using it. Anyone can add a course.
Honest about what this is
Cairn is a free side project. It exists because the person who made it kept abandoning Stanford CS and MIT OCW courses around lecture four, and wanted a tracker tool that is accessible. If that's you too — welcome.