Twitch clip captions: edit text and timing, burn them in
Twitch is rolling out editable clip captions. For a clip file you edit yourself, Sume burns your own cues with text, start and end, no speech-to-text.

Twitch's own caption editor works inside Twitch. If you have a clip as a video file and want the words, timing and look under your control, burn them with POST /v1/video-captions: pass cues with text, start and end and Sume draws exactly that copy, with no speech-to-text.
Twitch details are from its TwitchCon Rotterdam post, and Sume details from Video captions, read 2026-09-30.
What is Twitch adding for clip captions?
The post says Auto Clips generates captioned clips from the best moments of a stream. It also says Twitch is rolling out the ability to add captions to community clips automatically, with editable text, timing and style so they work for viewers watching with sound off.
What can Sume do with a clip file?
| Goal | Request |
|---|---|
| Burn your own wording and timing | cues (or segments) with text, start, end |
| Let speech-to-text time the words | Omit cues; optional language hint |
| Keep STT timing, fix the wording | script_text aligned to STT timings |
| Change the look | style, and optional design overrides |
| Re-burn in another style without transcribing twice | source_caption_id instead of video_url |
How do I get an editable transcript first?
Import the clip, then call POST /v1/video-inspect with transcribe: true and sentence segmentation to get lines with start and end. Fix the text in your own editor and send the result as cues. script_text, words, cues and segments are mutually exclusive, so send one. A silent clip fails speech-to-captions, and cues is the documented way around that.
Should I trim the highlight first?
If the moment you want sits inside a longer recording, cut it with POST /v1/video-trim using start plus end or duration (exactly one of the two), then caption the new file. Cue times are in seconds from the start of the video you caption, so caption after trimming. See burn captions onto a video for the request shape.
Sources
Related posts
More in Use cases
- Twitch vertical clips: crop an older 16:9 clip to 9:16
Twitch Dual Format makes vertical clips going forward. For a 16:9 clip without one, crop a 9:16 slice with video-filter: width 0.316, height 1.
- Walmart main image size: 2200x2200 and Sume's multiples of 16
Walmart wants a 2200x2200 px main image under 5 MB. 2200 is not a multiple of 16, so generate 2208x2208 on Sume's gpt-image-2.5, then resize.
- WordPress 7.1 client-side media processing and Sume images
WordPress 7.1 resizes and converts uploads in the browser (default quality 0.82). Sume has no output_compression, so request the format you want.
- AI-made captions on YouTube: is a disclosure label required?
YouTube lists caption creation among edits that need no altered-content disclosure. What that covers, and how a Sume burned-in caption job fits.
Written by Sume