Restyle captions three times with source_caption_id: $0.60 total

To try a second caption style on a video that Sume has already captioned, send source_caption_id instead of video_url. Sume reuses that caption's source video and the word timings it already holds, so speech-to-text does not run again. The price does not change, because a restyle is still a render. Three looks on one clip is three jobs at $0.20 each, $0.60 in total, under the docs' current estimate for clips up to 60 seconds.

The arithmetic

The caption docs state each accepted standalone job reserves and captures $0.20, and that a restyle is priced like any other job. The live price is in GET /v1/catalog.

What restyling saves

The saving is not money; it is consistency and time. The transcript and timings are fixed after the first job, so the three renders differ only in look. If a word is wrong, send words along with source_caption_id to correct the text.

The same applies to a typo you fix once: correct it in the restyle request, then keep the corrected job as your base for later restyles.

Request

Replace the id with the one from your first caption job.

curl -X POST https://api.sume.com/v1/video-captions \
  -H "Authorization: Bearer $SUME_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: restyle-clip-wipe-001" \
  -d '{
    "source_caption_id": "vc_123",
    "style": "clip-wipe"
  }'

Picking between Hangul looks

For Korean speech the docs describe clip-wipe as the clearest style at small phone sizes, and black-outline as the safe default. highlight moves an accent block behind the spoken word and pill-karaoke changes color inside a dark pill. Compare them on one real clip, because the way a style reads depends on the footage behind it.

When not to restyle

Restyle works from a finished caption job. If the transcript itself is wrong, change the words rather than the style. If the clip changed, upload the new video as a new job, because source_caption_id reuses the original source video and its timings.

Keep each caption job id next to the video it came from. Then a later brand refresh costs $0.20 per clip for a render, with no re-transcription and no new upload. Forty clips refreshed would be 40 x $0.20 = $8.00 under the same estimate.

• Standalone caption job: $0.20 under the current estimate for clips up to 60 seconds.
• Inline captions on an Avatar Video are a separate add-on and create no caption resource, so they cannot be restyled this way.
• Confirm the live rate in GET /v1/catalog before a large batch.