Slayshot is shut down.The work lives on.

Proof of what we built

Slayshot is no longer running. Before it went offline, it turned long videos into sharp, captioned clips ready to post. Here is some of the work it made.

Explore the code See the proof

Proof of work

What Slayshot made.
Before the shutdown.

Hooks found, faces tracked, captions timed, and layouts reframed for a screen held in one hand.

Made with Slayshot

Slayshot cut 01

Selected cuts

01 / 09

Real clips, cut from long-form video.

How it worked

One complete
clip pipeline.
Now open to inspect.

The hosted product is gone, but the pipeline is still here. Trace every step, swap any model, and deploy it wherever you want.

Find the signal

Transcription and scoring surface the moments with enough context to stand on their own.

Shape the cut

Turn the strongest sections into tight clips without dragging a timeline frame by frame.

Keep people in frame

Active-speaker detection reframes wide video into a vertical composition that follows the conversation.

Ship it captioned

Word-timed subtitles make every clip understandable with the sound off and ready for the feed.

The product is gone. The source is not.

Take what worked.
Build what is next.

Study the complete pipeline
Run it on your infrastructure
Fork it and take it further

Get the source