Seedance 2.5 Video Generation

Generate native 30-second video with up to 50 multimodal references, 4K output, and region-level editing through the PixMind API.

Native 30-second single-shot video with up to 50 multimodal references, 4K output, and region-level editing.

Provider

ByteDance

Model ID

seedance-2.5

Endpoint

POST /v1/generations

Open model playground Create API key

Model notes

Single-shot clips can run up to 30 seconds; longer sequences still need concatenation.
Up to 50 multimodal reference inputs (image, video, text, audio) can guide characters, scenes, style, and sound together.
Pricing shown is estimated and pending the official Seedance 2.5 release.

Request parameters

Parameter	Type	Required	Description
`model`	string	Yes	Stable PixMind model ID. Use `seedance-2.5`. Allowed values: `seedance-2.5`
`prompt`	string	Yes	Natural-language generation or editing instructions.
`resolution`	string	No	Output video resolution. This field affects per-second pricing. Allowed values: `480p`, `720p`, `1080p`, `4K` Default: `480p`
`duration`	integer	No	Requested clip duration in seconds. Allowed values: `5`, `10`, `15`, `20`, `30` Default: `5`
`aspect_ratio`	string	No	Requested output aspect ratio. Allowed values: `16:9`, `9:16`, `1:1`, `4:3`, `3:4` Default: `16:9`
`image_url`	string (URI)	No	Starting image for image-to-video generation.
`callback_url`	string (URI)	No	HTTPS endpoint that receives the terminal task event. Polling remains available when omitted.
`reference_images`	string[] (URI)	No	Reference image URLs used for editing, composition, subject, or style guidance.
`reference_videos`	string[] (URI)	No	Reference video URLs for character, scene, motion, or style guidance.
`generate_audio`	boolean	No	Generate synchronized audio when supported by the selected mode. Allowed values: `true`, `false` Default: `true`
`enable_web_search`	boolean	No	Allow supported modes to use web context during prompt interpretation. Allowed values: `true`, `false` Default: `false`

Request example

cURL

curl https://aihub-admin.aimix.pro/v1/generations \
  -H "Authorization: Bearer $PIXMIND_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "seedance-2.5",
  "prompt": "A 20-second cinematic product reveal with consistent character identity and synchronized ambience",
  "duration": 20,
  "resolution": "1080p",
  "aspect_ratio": "16:9",
  "reference_images": [
    "https://example.com/product.jpg",
    "https://example.com/character.jpg"
  ],
  "generate_audio": true
}'

JavaScript

const response = await fetch("https://aihub-admin.aimix.pro/v1/generations", {
  method: "POST",
  headers: {
    "Authorization": `Bearer ${process.env.PIXMIND_API_KEY}`,
    "Content-Type": "application/json"
  },
  body: JSON.stringify({
  "model": "seedance-2.5",
  "prompt": "A 20-second cinematic product reveal with consistent character identity and synchronized ambience",
  "duration": 20,
  "resolution": "1080p",
  "aspect_ratio": "16:9",
  "reference_images": [
    "https://example.com/product.jpg",
    "https://example.com/character.jpg"
  ],
  "generate_audio": true
})
});

const task = await response.json();

Accepted task response

Generation is asynchronous. Save the id from the response and poll GET /v1/generations/{id} until it reaches a terminal status.

202 Accepted

{
  "id": "img_01JXYZ...",
  "model": "seedance-2.5",
  "status": "processing"
}

API pricing

Configuration	Price
480p	$0.072 / second (est.)
720p	$0.176 / second (est.)
1080p	$0.336 / second (est.)
4K	$0.840 / second (est.)

API pricing uses the apiPlatform role and is separate from PixMind Studio credits. The accepted task response is the final pricing snapshot.