Clipnote – Private AI Video Summary & Notes

On-device AI tells you what any video is about, its key points, conclusion and chapters. Transcribes videos without captions.

Know what any video is about before you watch it — with an AI that runs on your own computer. Clipnote reads the video’s captions (or listens to it, if there are none) and tells you: • What this video is — in one sentence • Key points — the 3 to 5 things you need to know • Conclusion — what the video ultimately wants to say • Chapters — click a time to jump right there Also • Notes — write a note at any moment; each keeps its timestamp. Copy or download as Markdown for Notion or Obsidian. • Transcript — search everything that was said. • Library — every video you summarized is kept on your PC with its captions, transcript, AI summary and notes. Search them all at once (“which video said that?”) and jump to the exact moment. • Listen & transcribe — no captions? Clipnote listens while the video plays and writes it down. It detects the spoken language by itself and hears only the video you chose, not other tabs. • Video files on your PC — open meeting recordings, lecture recordings or interviews (mp4, webm, mov…) plus subtitle files. Nothing is uploaded. • Offline mode & network report — after the one-time model download, block every outside request. The panel shows the engine’s request count so you can see nothing was sent. Why Clipnote • 100% on-device: summaries (Google’s open Gemma model) and transcription (Whisper) run on your computer. Nothing is uploaded. No servers. • Free and unlimited. No account, no sign-in, no daily limits. • Works on YouTube automatically and on other video sites (courses, news, Wikimedia and more) with one click on the toolbar button. • Summaries in English or Korean; captions and speech in many languages. Requirements: a computer with a graphics card that supports WebGPU (most PCs from the last few years) for the AI summary. The models download once (about 760 MB for the summary, 80 MB for transcription). Without WebGPU, you still get key sentences picked from the captions. Protected streaming services (for example Netflix) can’t be transcribed. Clipnote is not affiliated with YouTube or Google. Gemma is provided under and subject to the Gemma Terms of Use found at ai.google.dev/gemma/terms.