Create lecture notes from video with a transcript and AI summary.

Drop a recorded lecture onto the box at the top of this page or choose a file. By default the audio is extracted in your browser and transcribed, then AI writes study notes — a summary, the key points, key terms with short definitions, and a timeline — that you can read on the page or download as Markdown, HTML, PDF or plain text.

Yes. Before you start you can set the spoken language of the recording and the language you want the notes in, separately — both default to auto-detect. So an English lecture can produce notes in your own language, while the transcript stays in the language that was actually spoken.

The full lecture transcript plus study notes: a summary of what was covered, the key points, a list of key terms with short definitions — useful when a lecture introduces a lot of new vocabulary — and a timeline so you can jump back to a topic to review it. Download the notes as Markdown (.md), HTML or PDF and the transcript as plain text (.txt).

Transcription quality follows the audio. If a recording is quiet, echoey, or heavily accented, setting the spoken language yourself instead of leaving it on auto-detect usually helps, because the model no longer has to guess which language it is hearing. Very noisy recordings will still produce mistakes — the full transcript is there so you can check anything that matters.

Yes. Common recordings work — MP4, MOV, WebM, MKV, plus audio files like MP3 and M4A. In Browser mode, the default, transcribe up to two hours per range, then move the range to the next part; Server mode (paid plans) processes the whole recording if your remaining audio time covers it. Browser mode, the default, takes files up to 2 GB; Server mode (paid plans) takes files up to 8 GB. YouTube lecture links work too, if the video has captions.