Seedance 2.0 Fast vs Seedance 2.0 (and Mini): what differs

Seedance 2.0 Fast and Seedance 2.0 take the same inputs and make clips of the same 4–15 second lengths; what you can measure between them is the price. Through Sume's API a 720p 16:9 second costs about $0.30 on Seedance 2.0 Fast and $0.38 on Seedance 2.0, and Seedance 2.0 Mini, the third member of the family, costs about $0.19.

Resolution is the one place the docs differ: they name 1080p for seedance-2 and don't mention it for Fast or Mini, so read each id's supported_resolutions before you ask for 1080p. The limits come from the Video generation docs and Sume's video catalog code; the prices are computed from the provider list prices in Sume's pricing code, read 2026-09-28.

What is the difference between Seedance 2.0, Fast and Mini?

In Sume's catalog code the three 2.0 ids share one capability block. Each does text-to-video, image-to-video with a first and a last frame, and reference-to-video with image, video and audio references, with optional sound. The docs confirm that audio and video references are honored by the Seedance 2.x models. They differ in the rate per 1,000 video tokens:

Is Seedance 2.0 Fast faster, or lower quality?

Sume's docs don't say. They publish no render times and no quality comparison between the three ids, and "Fast" and "Mini" are the model names, not documented measurements. What is documented is the list above: the same modes, the same lengths, and different rates.

If quality or wait time decides it for you, send the same prompt, length and resolution to each id and compare the clips and the usage.cost of each job. Only the model field changes between the requests.

Why does a Seedance second cost what it does?

Seedance bills video tokens, not seconds. Sume's pricing code counts them as width × height × seconds × 24 ÷ 1,024, which comes to 21,600 tokens for each second of a 720p 16:9 clip. Every 2.0 id makes the same number of tokens for the same clip, so the rate alone sets the price. Other aspect ratios and resolutions change the pixel count, and How much does AI video cost per second? compares those rates with other models.

Sume bills the provider's list price × 1.25, reserved from your workspace balance when you submit, plus a 5.5% agent fee by default. The poll response's usage.cost is the billed amount.

How is Seedance 2.5 different from Seedance 2.0?

Length is the documented difference: seedance-2.5 accepts 4–30 seconds at 480p, 720p or 1080p, while the 2.0 ids stop at 15 seconds. It also costs more, with a higher token rate at 1080p ($0.02925 per 1,000 tokens). Seedance 2.5 API: 30-second clips covers its limits and a 30-second request.

How do I switch between the Seedance ids?

Change model. The request shape is the same for every Seedance id on POST /v1/videos; GET /v1/videos/models lists each id's supported_resolutions, durations and pricing_skus before you call it. Send resolution explicitly.

curl -X POST "https://api.sume.com/v1/videos" \
  -H "Authorization: Bearer $SUME_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "seedance-2-fast",
    "prompt": "A golden retriever playing fetch on a sunny beach",
    "duration": 10,
    "resolution": "720p",
    "aspect_ratio": "16:9"
  }'