20 s Seedance: only 2.5 passes 15 s, $11.56 vs two 2.0 jobs at $7.56

For one 20-second Seedance clip, you need Seedance 2.5. On Sume, 2.5 accepts 4 to 30 seconds, while Seedance 2.0, Fast and Mini stop at 15. At 720p and 9:16 the 20-second 2.5 job costs $11.56. Two 10-second Seedance 2.0 jobs cost $7.56 together, but they are two separate clips, so continuity is yours to manage.

The numbers

Seedance rows are priced by video tokens, and Sume bills the list rate times 1.25, rounded up to the cent. The grid below applies that to 720p 9:16.

What the lower price costs you

The gap between one 2.5 job and two 2.0 jobs is $4.00. For that you get a single continuous shot instead of a cut at 10 s, and a model that accepts the full 30-second range. Two 2.0 jobs need the second job to start where the first stopped. The catalog supports last-frame-as-first-frame handoff through frame_images, but the second clip is a new generation, so subtle drift can appear at the join.

Limits worth knowing

• Seedance 2.5, 2.0, Fast and Mini share the same ratio list (21:9 through 9:16, plus auto and adaptive) and the same 480p, 720p and 1080p set.
• 2.5 has no bitrate_mode.
• Wan 3.0 also reaches 30 s, at a much lower price, but it lists no 21:9 ratio.

How to choose

If the shot must be one take and Seedance 2.5 is the look you want, pay for 2.5. If a cut at 10 s is acceptable, two 2.0 jobs save $4.00. If Seedance is not a requirement, a single 20-second Wan 3.0 job at $2.50 is the cheapest route to 20 seconds. Submit through Video generation with the pinned model id, and read supported_durations first.

Check it before you run it

Every figure above is a catalog list price times 1.25, rounded up to the cent, as of 2026-10-08. Catalogs change, so before a large batch, read the current model entry in the docs and recompute the one line that matters for your case. Write the arithmetic next to the job in your own notes: list rate, seconds or characters, multiplier, rounding. If the result differs from the wallet charge by more than a cent, the catalog entry has changed, and the docs page is the place to find out why.

Run one small job first. Submit a single request with the pinned model id and the settings in the tables, poll the returned polling_url until it finishes, and compare the charge with your estimate. Then scale up. Using a pinned id for the test matters, because sume/auto never names the family, so you cannot tie its charge to the row you priced.