Seedance 2.5

Seedance 2.5 (bytedance/seedance-2.5) is the ByteDance video model built for long-form storytelling: multimodal references, first-plus-last frame support, and native audio. Generation is asynchronous: POST /v1/videos returns a job, then you poll the status and download the mp4.

Parameters

Field	Description
model	bytedance/seedance-2.5 (or the short seedance-2.5)
prompt	Scene description, required
seconds	Any integer from 4 to 30; defaults to 4
size	A WxH size from the table below; defaults to 1280x720
input_reference	Image-to-video: one image as the starting frame, {"image_url": "data:image/...;base64,..."}
input_references	Reference-to-video: 2 to 30 images as an array, also data: URLs only — 2.5 takes far more references than its siblings
seed	Reproducibility seed: an integer in 0…2147483647. The same prompt with the same seed yields the same clip; without it the upstream picks a random one

Twelve allowed sizes — six aspect ratios at two resolutions:

Aspect	480p	720p
16:9	852x480	1280x720
9:16	480x852	720x1280
1:1	480x480	720x720
4:3	640x480	960x720
3:4	480x640	720x960
21:9	1120x480	1680x720

A value outside the list returns 400 with the allowed options listed — check the sizes field in the catalog when in doubt.

Examples

Text-to-video

Python

import requests

job = requests.post(
    "https://api.mixen.ai/v1/videos",
    headers={"Authorization": f"Bearer {API_KEY}"},
    json={
        "model": "bytedance/seedance-2.5",
        "prompt": "A drone circles a lighthouse on a rocky shore, dawn, spray from the waves",
        "seconds": "8",
        "size": "1280x720",
    },
).json()
print(job["id"], job["status"])

cURL

curl https://api.mixen.ai/v1/videos \
  -H "Authorization: Bearer $MIXEN_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "bytedance/seedance-2.5",
    "prompt": "A drone circles a lighthouse on a rocky shore, dawn, spray from the waves",
    "seconds": "8",
    "size": "1280x720"
  }'

Image-to-video from a starting frame

The image goes in as a single data: URL — external links are not accepted:

{
  "model": "bytedance/seedance-2.5",
  "prompt": "The camera pulls back slowly, the city wakes up, natural motion",
  "seconds": "10",
  "size": "720x1280",
  "input_reference": {"image_url": "data:image/jpeg;base64,/9j/4AAQ..."}
}

Reference-to-video from several images

Two to nine references: the model assembles the scene while keeping the characters and style of the inputs:

{
  "model": "bytedance/seedance-2.5",
  "prompt": "The girl from the first frame walks down the street from the second, evening lights",
  "seconds": "6",
  "size": "1280x720",
  "input_references": [
    "data:image/png;base64,iVBORw0...",
    "data:image/png;base64,iVBORw0..."
  ]
}

From there it is the usual loop: GET /v1/videos/{id} every 5–10 seconds until completed, then GET /v1/videos/{id}/content for the mp4. Full cycle — video guide.

Price

Billing is per second and depends on the resolution: 480p — 13.7436 ₽ ($0.1542) per second, 720p — 30.8991 ₽ ($0.3467). You are charged actual duration × the size’s rate. Current prices — catalog.

Strengths and limits

• Long-form stories: durations from 4 to 30 seconds — a continuous range, not three fixed buttons.
• Six frame formats, including the cinematic 21:9, at 480p and 720p.
• Native audio in the finished clip.
• Image-to-video from a starting frame and reference-to-video from 2–9 images.
• The first-to-last frame transition is available in the web cabinet and the bot; over /v1 only the starting frame can be passed.
• 720p costs roughly twice as much as 480p — on a 30-second clip the difference adds up.

All models — in the catalog. The general video workflow — video guide.