Part of ScenePilot, from thecreateco
Browser tool / PUBLIC WEB ALPHA
Captions
Say it once. Caption it everywhere.
Captions is ScenePilot's on-device speech-to-text tool: drop in video or audio and get an editable, word-timed transcript, five animated caption styles, and a burned-in WebM export alongside SRT/VTT files you can take anywhere.
Browser tool · on-device captions, not a translation or dubbing service
Public alpha · Availability & current scope · Send feedback
WHAT IT DOES
Turn a voice track into styled, exportable captions.
Drop in a clip and Captions runs an on-device speech model to build a word-timed transcript you can edit without breaking the timing — fix a word, change the word count per line, or merge and delete lines. Pick from five animated styles, then burn the captions into a WebM export or take SRT/VTT files elsewhere. Audio-only clips get a 9:16, 1:1 or 16:9 audiogram backdrop, and you can also import and restyle an existing SRT/VTT.
HOW IT WORKS
Three steps to get there.
- 01
Drop in a video or audio clip.
Captions runs on-device speech-to-text and builds a word-timed transcript automatically.
- 02
Edit the transcript and pick a style.
Fix a word, regroup lines with the caption-length slider, and choose one of five animated presets.
- 03
Burn in a WebM or export SRT/VTT.
Keep the styled video, or take the caption files into ScenePilot or another editor.
HONEST LIMITS
What Captions doesn't do.
Current limitations
Captions uses Whisper tiny/base on-device, so names and sung lyrics often need manual fixes. Burn-in renders in real time and the browser tab needs to stay visible. There's no MP4 export yet — WebM only — and word-by-word animation only exists in the burned-in video; handing a cut off to the ScenePilot editor carries over per-line titles only. This is a public web alpha; output quality is still changing.
One good first test
- 01Drop in a short clip with clear speech.
- 02Try a couple of animated caption styles.
- 03Burn in a WebM or export the SRT file.
MORE OF SCENEPILOT