Captions for 100 ad variants: transcripts at one cent per audio minute

Transcripts on Sume are priced at $0.01 per audio minute. A batch of 100 ad variants at 15 seconds each is 25 audio minutes in total, which is $0.25 at that rate, assuming you plan on the audio length. Real jobs may round differently, so treat it as a planning figure.

This matters for holiday catalogs where each variant needs caption text for the burned-in line. Transcripts are a rounding error next to the video render.

The numbers

Audio minutes = variants x seconds / 60. Cost = minutes x $0.01. The catalog lists the speech-to-text SKU and the video-inspect transcript SKU at the same $0.01 per audio minute.

Where captions fit

A transcript gives you the words and their order. You can use that to place a price line or a countdown cue at the right moment. The Black Friday price variants post shows one clip with 12 caption cues.

If the audio comes from a voiceover you generated, you already have the text. Use the transcript as a check that the spoken price matches the script.

Compared with the render

The transcript for 100 variants is $0.25. The clips themselves at 6 seconds, 480p, 9:16 on Seedance 2.5 are 100 x $1.612063 = $161.2063. Captions are about 0.16 percent of that.

So the question is workflow, not cost: where in your pipeline do the transcript and the review sit.

Practical notes

• Transcript jobs hold and settle like other jobs. See errors and credits.
• Round up your audio length when you plan, since real files are rarely exactly 15 seconds.
• Check caption accuracy yourself, especially for prices and dates.