Drop The Rush. Get The Clips.
For creators who want to film once and post ten times.
An AI clipper for creatorsUpload a long recording; the engine finds the moments worth posting and cuts them to 9:16.
What already runs today.
The engine is built and used on our own long-form sources. Four steps, the model decides what to cut, the renderer never improvises.
Ingest
A YouTube link or a local recording. The source is probed once, whatever its length or orientation.
Transcribe
A word-timed transcript, from the platform's own captions when they exist, from a local speech model otherwise.
Detect
The whole transcript goes to the model, which returns the moments worth a clip, ranked, each with its cut points and a hook. You approve.
Render
9:16, the original audio, a hook card on top and a part number under the video. One deterministic pass per clip.
What the app adds.
The waitlist gets the first builds and sets the order of what ships. No price yet: it will be decided with the people on it.
For creators and the people who edit for them.
- Upload from the browser instead of the terminal
- Review the proposed moments, retime a cut, rewrite a hook
- Captions burned in, with a style per account
- Publish to the same connected accounts as Farmiz Slideshows
Film the messy take.
Post the clean one.
Record once. The engine screens the recording, keeps the moments worth posting and hands back the clips.
Join the waitlist