Finishing a 30-episode series: captions, upscale, cutouts on Sume

The finishing tools for a 30-episode series of 45-second episodes are cheap next to generation: captions from speech-to-text cost $0.23 for all 22.5 minutes, 30 cover cutouts $0.68 and 30 cover upscales $6.00. Upscaling every second of video is the one large line at $12.15. The five-and-a-half percent platform fee applies to each.

Where the series comes from

TechCrunch reported on 2026-09-23 that YouTube Shorts series support seasons, episodes and custom thumbnails. Thirty episodes of 45 seconds is 1,350 seconds, or 22.5 minutes; the 45-second length is this post's assumption, matching the series budget post.

Prices used

These are Sume's published API rates: speech-to-text $0.01 per minute, background removal $0.0225 per image, image upscale $0.20 per image and video upscale $0.009 per second. The speech-to-text line assumes minutes are billed pro rata, which is how this arithmetic treats them; per-job rounding would change it slightly.

Optional steps

All four steps together are $19.05 before the fee and $20.10 after it, and $12.15 of that is the video upscale. Skip the video upscale if you generate at the resolution you publish at, and the remaining three come to $6.90 before the fee. Whether you need any of them depends on the models and resolutions you used.

Running them

Each step is its own job with its own reserve and refund, so a failed upscale does not stay billed. The Sume pricing page lists the rates; confirm them before queuing a batch.