Runway API Seedance: calling Seedance 2.5 and 2.0 on Runway
Yes, Runway's API serves Seedance: seedance2_5, seedance2, seedance2_fast and seedance2_mini. Inputs, limits and credits, and other ways to call Seedance.

Yes: Runway's API serves Seedance. Its model list includes seedance2_5, seedance2, seedance2_fast and seedance2_mini, each taking text, image or video input, on the same task endpoints as Runway's own models. You pick one with the model field and pay in Runway credits.
Every Runway fact below comes from Runway's own developer docs, read on 2026-09-28 and listed under Sources. Does Runway have an API? covers keys, SDKs and polling; this page is only about Seedance.
Which Seedance models does the Runway API list?
Runway's models page lists four Seedance ids, each taking text, image or video input. Their request limits differ, so a body written for one is not valid for another. Runway bills them in credits per second of output, at $0.01 a credit, with a rate per model and resolution; Runway API pricing per second lists the rates.
| Model id | Prompt | `ratio` | References |
|---|---|---|---|
seedance2_5 | Optional, up to 15,000 characters | 18 values, up to 1080p | 30 images; 10 videos, 30 s combined; 10 audio clips, under 30 s |
seedance2 | Required, up to 3,500 characters | 24 values, up to 4K | 9 images; 3 videos, 15 s combined; 3 audio clips, 15 s |
seedance2_fast | Required, up to 3,500 characters | 12 values, up to 1280:720 | 9 images; 3 videos, 15 s combined; 3 audio clips, 15 s |
seedance2_mini | Required, up to 3,500 characters | 12 values, up to 1280:720 | 9 images; 3 videos, 15 s combined; 3 audio clips, 15 s |
What does Seedance 2.5 accept on Runway?
On POST /v1/text_to_video, seedance2_5 takes an optional prompt of up to 15,000 characters, audio (on by default), a seed, a ratio (“Seedance 2.5 supports 480p, 720p, and 1080p”), and a duration in seconds or auto, plus the references in the table (API reference).
POST /v1/image_to_videotakespromptImagewith positionfirstorlastfor keyframes, or no position for reference images; the two modes can't be mixed.POST /v1/video_to_videotakespromptVideoand amodeofreference(the default),extendoredit.duration: "auto"is billed at the variant maximum up front, and unused credits are refunded after the generation finishes.
What about Seedance 2.0 on Runway?
Runway's primer gives seedance2 on POST /v1/text_to_video a duration from 4 to 15, 24 ratios up to 4K and a 3,500-character prompt, with audio and all three reference types (Runway Dev context). It is the only Seedance id there with 4K ratios; seedance2_fast and seedance2_mini stop at 1280:720 (API reference).
Where else can I call Seedance by API?
BytePlus's Seedance page says the Seedance 2.5 API is “fully available now” (BytePlus); Sume vs Seedance covers that route. Multi-model APIs serve it too.
On Sume, Seedance runs on POST /v1/videos with bare ids. seedance-2.5 accepts 4–30 seconds at 480p, 720p or 1080p, and seedance-2 4–15 seconds with 1080p; the Seedance 2.x models honor image, video and audio references, and Seedance is priced per 1,000 video tokens (Video generation). Rates are on API pricing, and Seedance 2.5 API lists every Seedance id.
Seedance or Runway: which should I use?
It isn't either-or: on Runway's API, Seedance is one of the models you can pick. Choose by what the rest of the job needs:
- Runway if you also want Runway's own models, its Model Routers, or Seedance 2.0 at 4K, which its pricing lists.
- BytePlus if you only want Seedance, from the page that sells it directly.
- A multi-model API such as Sume if Seedance is one of several video models your code calls on one balance.
- Whichever route, the ids, limits and billing units differ, so don't reuse a request body across them.
Sources
Related posts
More in Models
- Runway Gen-4.5 vs Seedance 2.0: inputs, length and price
On Runway's API, Gen-4.5 makes 2–10 s clips for 12 credits a second; Seedance 2.0 makes 4–15 s clips with audio for 36–150 credits a second.
- Runway ML alternatives for AI video: apps and APIs
Runway ML alternatives for AI video: fal, Higgsfield, Kling, Replicate and Sume compared on app vs API, models and billing, read 2026-09-28.
- Seedance 2.0 vs Gemini Omni Flash: inputs, audio and price
Seedance 2.0 makes 4–15 s clips with optional audio and audio references; Gemini Omni Flash 1.1 makes 3–10 s clips up to 4K, always with sound.
- Seedance 2.5 vs MiniMax H3: length, resolution, audio, price
On Sume, Seedance 2.5 makes 4–30 s clips at up to 1080p with optional audio; MiniMax H3 makes 5–15 s at native 480p or 768p, always with audio.
Written by Sume