Break down
Studying one minute of video properly takes half an hour. Beav does the cutting, transcribing and screenshotting first, so you can focus on why it was shot that way.
Desktop app for macOS, Windows and Linux
For talking-head, food, unboxing and tutorial videos, see what each shot shows, says and how long it lasts. For stories and visits, watch the clips; for talking heads, read the transcript next to the shot script.
Multi-select in the knowledge base and right-click to decompose the selected items, up to 100 at once. Queued, running and finished items are skipped. Dozens in one format show which openings keep recurring.
Click + and Add link, then paste a RedNote, Douyin, Bilibili or other video link. The video must be downloaded to your computer before shots and clips can be cut; wait while the card says Downloading.
You get five parts: topic, structure template, shots, shot-by-shot script and timestamped transcript. The script says what each shot shows and how; every transcript line has a time to match the picture.
Click a shot thumbnail or time to jump there. If two shots are really one, merge it into the previous shot; if one shot hides a cut, play to it and split at the playback position. Decomposing again keeps your cuts.
The decompositions/<title> folder holds the script, keyframes, clip MP4s and music M4A, with the source video, creator and clip times recorded. The topic is added to the topic center.
What each shot shows and how, including framing and angle changes.
One clip and one keyframe per shot, in the asset library.
Every line timed to line up with the shots.
How it opens, turns and closes.
| Aspect | Studying by hand | Asking a general AI (Gemini, ChatGPT…) | Beav Decompose |
|---|---|---|---|
| What you get | Your notes and screenshots | A text analysis to follow up on | Topic, structure, shot script, timestamped transcript |
| Clips and keyframes | Cut and capture yourself | Usually text only | Clip MP4s and keyframes saved automatically |
| Adjusting cuts | Fully yours | Describe and ask again | Merge into the previous shot or split at the playback position |
| How many at once | One | One | Up to 100 selected |
| Cost | Time | Their memberships | Model quota |
Studying by hand trains your own eye, which no tool replaces; a general AI is good for open-ended questions about one video. Beav works like a breakdown assistant: it does the repetitive work and the results become material in your library.
Save the video to the Beav knowledge base, let it download, then click Decompose for a shot-by-shot script, shot list and timestamped transcript.
Yes. Multi-select in the knowledge base, right-click to decompose the selected items, confirm the count, duration and model cost, up to 100 at a time.
No. Decomposing only runs when you click it, so you control the cost.
Decomposing uses transcription, chat and video-understanding models and consumes model quota. Batches show the cost first.
Videos above the video-understanding size limit get a low-resolution copy first; if still too large, the shot script is skipped with a reason and the rest is saved.
They are for study and reference. Confirm you have the rights before publishing or commercial use.
Break down
Keep a video's music as M4A, with the source attached.
Break down
Paste many Douyin links and pull the spoken scripts in one batch.
Create
Keep the original's structure, switch to your language and product.
