Same image as first and last frame: a 6-second loop on three models

Send the same image as first_frame and last_frame and the clip starts and ends on that picture. A 6-second loop costs $2.268 on Seedance 2.0 at 720p, $0.60 on MiniMax H3 Max at 768p and $0.75 on Gemini Omni Flash 1.1 at 720p. The docs name these three as taking an end frame.

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.

Which models take a last frame

Seedance 2.0 lists supported_frame_images of first_frame and last_frame. H3 Max does start/end-frame image-to-video. Omni takes image_url plus an optional end_image_url. For any other model (including Wan 3.0 and Seedance 2.5) read supported_frame_images on GET /v1/videos/models before you rely on it, since this page only repeats what the docs state.

Loop quality

A loop needs the motion to return, and a model may drift in the middle even when both ends match. A shorter clip drifts less, so for a loop use the minimum that fits the motion: 4 s on Seedance 2.0 is $1.512, 5 s on H3 Max at 768p is $0.50, 3 s on Omni at 720p is $0.375.

If you send both frame_images and input_references, frame_images controls the mode and the request runs as image-to-video.

Scaling this plan up or down

As a yardstick, the plan above is built on Seedance 2.0 at 720p, $0.378 per second. Each extra 5 seconds adds $1.89, a further $10 of budget buys 26 more whole seconds, and the largest single job the model accepts (15 s) holds $5.67 at submit. The shortest one (4 s) holds $1.512.

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: $0.378
• Per 5 s: $1.89
• Per 10 s: $3.78
• Per 30 s or the model maximum (15 s): $5.67

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.