A 3-second beat on Seedance 2.5 is a 4-second job: $2.31 at 720p
Seedance 2.5 starts at 4 s, so a 3-second beat is a 4 s job: $1.07 at 480p, $2.31 at 720p, $5.69 at 1080p, plus $0.02 if you trim it back to 3 s.

A 3-second beat on Seedance 2.5 is a 4-second job: $1.074708 at 480p, $2.3112 at 720p, $5.6862 at 1080p. If the edit needs exactly 3 seconds, add the $0.02 video trim job, which makes the 720p beat $2.3312.
Prices on this page are Sume list prices as of 2026-10-08: the provider list price times 1.25, computed from Sume's pricing package and billed per output second, linear in duration inside each model's valid range. The model ids, ranges and input types come from the Video generation docs and the Video Router docs; the per-second numbers can be cross-checked in the public catalog.
The cost of the extra second
The floor is one of the first things to check when you budget short clips, because it makes the shortest job cost four seconds of the model's rate whatever length you wanted. The extra second is billed at the same rate; nothing is discounted for the part you cut away.
| Resolution | Rate per second | 4 s job | Same beat at 3 s (not available) | Extra paid for the unused second |
|---|---|---|---|---|
| 480p | $0.268677 | $1.074708 | $0.806031 | $0.268677 |
| 720p | $0.5778 | $2.3112 | $1.7334 | $0.5778 |
| 1080p | $1.42155 | $5.6862 | $4.26465 | $1.42155 |
Cheaper ways to get a 3-second beat
Gemini Omni Flash 1.1 accepts 3 s directly: 3 s × $0.125 = $0.375 at 720p with native audio always on. Wan 3.0 accepts 2 s and up: 3 s × $0.125 = $0.375 at 720p.
If the look must come from Seedance 2.5, generate once at 4 s and keep the spare second as handles for the cut. Handles are useful for transitions, so the extra second is not always waste.
- Omni at 720p, 3 s: $0.375
- Wan 3.0 at 720p, 3 s: $0.375
- Seedance 2.5 at 720p, 4 s: $2.3112
Scaling this plan up or down
As a yardstick, the plan above is built on Seedance 2.5 at 720p, $0.5778 per second. Each extra 5 seconds adds $2.889, a further $10 of budget buys 17 more whole seconds, and the largest single job the model accepts (30 s) holds $17.334 at submit. The shortest one (4 s) holds $2.3112.
Those three numbers are enough to rescale the plan without a new table. If the plan doubles, double the totals; if a clip is shortened, subtract the seconds multiplied by the rate; and if the tier changes, swap the rate for the one in the tables above.
- Per second: $0.5778
- Per 5 s: $2.889
- Per 10 s: $5.778
- Per 30 s or the model maximum (30 s): $17.334
Check the model's fields first
Put the price table in your own config and compute the total before you call the API. The price is linear in seconds inside a model's range, so a function of rate and duration is enough, and it lets you refuse a batch that goes over a ceiling. Re-read the catalog when you plan a large run, because list prices are what Sume bills from, and they can change.
Submitting
Submit through POST /v1/videos with model, prompt, duration in whole seconds and resolution; read GET /v1/videos/models first, because supported_durations and supported_resolutions differ per model. Sume reserves the price at submit and usage.cost on the poll response is the billable amount, so the figures here are what leaves the balance.
Sources
Related posts
More in Pricing
- 30, 60 and 90-second presenter videos: $5.52, $11.04, $16.56 Standard
On Sume, 30 s, 60 s and 90 s of avatar video cost $5.52, $11.04 and $16.56 at Standard, $7.35, $14.70 and $22.05 at Plus, $16.50, $33.00 and $49.50 at Max.
- 30 days of social images on GPT Image 2.5: 12 cents a day, $3.60
A month of one finished image per day, with 3 low drafts, 1 medium refine and 1 high final each day, costs 12 cents a day and $3.60 on Sume (GPT Image 2.5).
- 30-second 1080p AI video: Wan 3.0 $7.50 vs Seedance 2.5 $42.65
A 30-second 1080p clip costs $7.50 on Wan 3.0 and $42.65 on Seedance 2.5 through Sume's catalog. The full 480p, 720p and 1080p table, with the request.
- 30-second explainer on four video models: cost and how many fit $20
One project, six 5-second shots, three stills, 450 characters of voice, one music bed, one render: total per video model and how many fit a $20 wallet.
Written by Sume