Caption a 60-second video in one job: $0.20, not $0.60 in splits

Standalone captions on Sume cost a flat $0.20 per accepted job, for videos of up to 60 seconds. That makes one 60-second video one $0.20 job, and the same video cut into three 20-second parts three $0.20 jobs, or $0.60. Do not split a video that fits inside 60 seconds.

Per-job pricing, in numbers

The docs state that each accepted caption job reserves and captures $0.20 for a video of at most 60 seconds, under the current fixed estimate. The price does not scale with the length below that limit, so a 5-second clip and a 60-second clip cost the same.

Where flat pricing helps and hurts

Flat pricing helps long clips and hurts very short ones. A 6-second hook costs the same $0.20 as a 55-second explainer, so captioning 100 short hooks one by one costs $20.00. The longer clips carry the lower per-second cost.

If you can join short clips before captioning, the join is a separate step and bills separately. Timeline renders cost $0.10 per output minute, rounded up, so a 55-second render is $0.10. Joining ten 6-second hooks into a 60-second reel and then captioning it once costs $0.30 for the pair, instead of $2.00 for ten caption jobs. Whether that fits depends on whether you ship the reel or the single clips.

• Over 60 seconds, the standalone caption job is not the right size; split the video.
• Shipping the clips one by one needs one caption job each.
• Shipping a reel needs one render and one caption job.

Over 60 seconds

The $0.20 price is stated for videos of at most 60 seconds. A longer video is not a candidate for a single standalone caption job. Split it and the number of caption jobs is the number of parts.

A 150-second video split into three parts of 50 seconds needs three jobs, or $0.60. A 130-second video split into 60, 60 and 10 seconds needs three jobs as well. Two parts of 65 seconds would not fit, so plan the cut points to land under 60 seconds.

Use the video-trim endpoint for the cut; it bills $0.02 per job, so three cuts add $0.06 to the caption cost.

Checking the figure

Before you queue a caption batch, read the price from GET /v1/catalog rather than from a cached note, since the docs name the catalog as the live source. The fixed estimate can change; the logic of one job per video does not. Multiply the live price by the number of videos, then compare the result with GET /v1/balance.

Check the number

Confirm the live price in GET /v1/catalog, which the caption docs point to. If the price changes, the table here changes with it: the rule is the flat per-job figure times the number of jobs. For the inline add-on on avatar videos, see inline captions versus a standalone job, and what it costs to finish 100 holiday ads for the whole finishing line.