Format DNA from a reference video
Paste a video whose format you like. VidSwarm reads how it is built and makes a format from it, without copying the video.
Beta
How to use it
Paste a YouTube URL into the format field on /app/c/[channel]/videos/new, or on /app/formats under “From a reference”. Reading takes a minute or two. The result is a format card, the same as for a described format, marked with the reference it came from.
What is read
- Structure: how the hook works, where chapters turn, how long they run, how the film ends.
- Pacing: average scene length and how it changes across the film.
- Scene types: footage, maps, charts, quotes, text, reconstructions, and how often each appears.
- Graphic language: the kinds of motion graphics used and how text is placed.
- Narration: speed, sentence length and register.
What is not taken
Format DNA describes how a video is built. It does not reuse the reference’s script, footage, music, graphics or thumbnail, and films made from it render in your channel’s look, not the reference’s. Copying another channel closely is both a copyright risk and a monetization risk under YouTube’s inauthentic content rule.
Combining references
You can read two or three references into one format. VidSwarm shows where they differ, for example one fast and one slow, and asks which to follow. Mixing two channels’ structures is often a good way to arrive at something that is clearly yours.
Limits
Very short videos, Shorts and videos without narration give little to read. Private and age-restricted videos cannot be read. For best results, use a long-form video of at least five minutes in the language you plan to produce in.