Break down a video
Turn a video into a study breakdown
Drop the video or audio file itself (mp4, mov, webm, mp3, m4a, wav…) and Novus Learn transcribes it on your device with a one-time speech-model download — or, for an exact time-coded breakdown, upload a caption file (.vtt or .srt) or paste the transcript. Everything runs in your browser; nothing is uploaded, and nothing is invented.
Drop a caption file here, or choose one
Captions: .vtt, .srt.
On-device transcription isn’t active on this origin yet. Captions (.vtt, .srt) and pasted transcripts still work. Video/audio (mp4, mov, webm, mp3, m4a, wav) needs the speech model served from this site — locally run npm run fetch:asr-model then restart npm run dev; on a deploy use the build:deploy pipeline. Then download the model once in the browser.
Captions, transcripts, and video/audio files
- Captions (.vtt or .srt) are the most accurate source — many courses and video players offer a “download captions/subtitles” option, and a “Show transcript” panel can be pasted above.
- No captions? Drop the video or audio file itself (mp4, mov, webm, mp3, m4a, wav…) and Novus Learn transcribes it on your device with an on-device speech model — you review and fix the draft before anything is built.
- Everything is processed in your browser and stored only on this device — nothing is uploaded.
- Every claim links back to the exact caption, timestamp, or line it came from.
What “break down a video” does
Watching a tutorial or a lecture is linear and slow to revise from. Break down a video turns its captions into a study space: a summary, structured notes split into clean modules, flashcards, a quiz, and a concept map — with every point linked back to the exact moment it was said.
The most accurate source is the video's own captions. Upload a caption file (.vtt or .srt) — exported from a course or a video's transcript panel — or paste the transcript. Novus Learn reads the captions verbatim, so the breakdown never invents anything the video did not say.
How it works
Get the captions
Download the video's caption/subtitle file (.vtt or .srt), or copy the text from its “show transcript” panel.
Upload or paste
Drop the caption file in, or paste the transcript. Everything is read and parsed in your browser — nothing is uploaded.
It becomes a source-grounded project
The captions are split into time-coded blocks and the same study tools are generated from them, each claim tied to its timestamp.
Study the modules
Work through the summary and notes as clean modules, drill the flashcards and quiz, and jump back to any moment in the source.
Why captions, not guesswork
Captions are the video's real words, so a breakdown built from them stays faithful to the source and every claim is checkable against the exact moment it came from — the same source-grounded guarantee as every other source in Novus Learn.
On-device transcription for videos without captions is a labelled, opt-in option (it downloads a speech model that runs entirely in your browser); captions remain the most accurate path.
Great for
- Tutorial and how-to videos broken into clear, ordered steps.
- Course lectures turned into notes, flashcards, and a quiz.
- Long talks distilled to their main points, each timestamped.
Frequently asked questions
What do I upload?
A caption file (.vtt or .srt), or a pasted transcript. These are the video's real words, so the breakdown is accurate and time-coded.
Can I break down a video without captions?
On-device transcription (a speech model that downloads and runs entirely in your browser) is a labelled, opt-in option. When captions exist, they are the most accurate source and add nothing that was not said.
Is anything uploaded to a server?
No. Captions and transcripts are parsed in your browser and the project is saved on this device only.
Do the notes and quiz reflect the video exactly?
Yes. They are built from the caption text, and every claim links back to the moment it was said. Nothing is generated beyond what the captions contain.