Reach APIReach API
Dashboard
API ReferenceVideo APIsSeedance

Seedance 2 API

ReachAPI exposes the Seedance 2 family through the public model IDs seedance-2 and seedance-2-fast.

This page describes how to call Seedance 2 models on ReachAPI, including request fields, constraints, and examples.

Important:

  • Create task: POST /v1/vids/create
  • Query task: GET /v1/tasks/{task_id}
  • Model IDs: seedance-2, seedance-2-fast
  • Execution mode: asynchronous task
  • All requests must now send explicit input.content

1. API Overview

  • HTTP method: POST
  • Request path: /v1/vids/create
  • Content-Type: application/json
  • Result delivery:
    • The submission response returns task acceptance information
    • The final result is obtained by polling /v1/tasks/{task_id} or by callback_url

2. Authentication and Headers

Example headers:

Authorization: Bearer YOUR_REACH_API_KEY
Content-Type: application/json

3. Model Capability Summary

ModelDurationResolutionAspect ratioinput.contentNotes
seedance-24-15s480p, 720p, 1080p21:9, 16:9, 4:3, 1:1, 3:4, 9:16, adaptiveRequiredSupports text, image, video, audio reference, editing, and extension
seedance-2-fast4-15s480p, 720p21:9, 16:9, 4:3, 1:1, 3:4, 9:16, adaptiveRequiredSame fields as seedance-2, but 1080p is not currently supported

Notes:

  • duration_seconds = -1 is supported and allows the model to choose the duration automatically

Billing and Usage

Seedance video tasks are billed by returned token usage, not by video seconds. Estimated price = token unit price × token consumption. The final bill uses usage.output_tokens in the ReachAPI task response, mapped from BytePlus usage.completion_tokens.

ModelVideo inputResolutionToken price
seedance-2No480p / 720p$7.00 / 1M tokens
seedance-2No1080p$7.70 / 1M tokens
seedance-2Yes480p / 720p$4.30 / 1M tokens
seedance-2Yes1080p$4.70 / 1M tokens
seedance-2-fastNo480p / 720p$5.60 / 1M tokens
seedance-2-fastYes480p / 720p$3.30 / 1M tokens
seedance-2-fastAny1080pNot supported

4. Create Task

Request body example:

{
  "model": "seedance-2",
  "callback_url": "https://your-domain.com/callback",
  "input": {
    "content": [
      {
        "type": "text",
        "text": "A cinematic drone shot passing over a coastline at sunrise"
      }
    ],
    "duration_seconds": 6,
    "resolution": "720p",
    "aspect_ratio": "16:9"
  }
}

4.1 Top-Level Parameters

FieldTypeRequiredDescription
modelstringYesMust be seedance-2 or seedance-2-fast
callback_urlstringNoHTTPS callback URL for terminal task states. Maximum length 2048. Localhost, local network, and private network targets are rejected
inputobjectYesVideo generation parameters

4.2 input Parameters

FieldTypeRequiredDescription
contentarray<object>YesExplicit multimodal input. Must be a non-empty array
duration_secondsintegerNoOutput duration. Use 4-15 or -1 for automatic duration selection
resolutionstringYesOutput resolution. seedance-2: 480p, 720p, 1080p; seedance-2-fast: 480p, 720p
aspect_ratiostringNoOutput aspect ratio. Supported values: 21:9, 16:9, 4:3, 1:1, 3:4, 9:16, adaptive
generate_audiobooleanNoRequests generated audio
seedintegerNoRandom seed. Supported range: -1 to 4294967295. Use -1 for a random result
safety_identifierstringNoEnd-user safety identifier. Must be an ASCII string up to 64 characters
watermarkbooleanNoPasses watermark preference to the underlying model
return_last_framebooleanNoRequests the last frame image
edit_modestringNoControls edit behavior in supported scenarios
extend_modestringNoControls extension behavior in supported scenarios

5. input.content Structure

Supported type values:

  • text
  • image_url
  • video_url
  • audio_url

Supported role values:

  • first_frame
  • last_frame
  • reference_image
  • reference_video
  • reference_audio

Field details:

FieldTypeRequiredDescription
content[].typestringYesMust be text, image_url, video_url, or audio_url
content[].rolestringNoIf provided, it must be one of the supported roles above. Non-text items should usually provide a role explicitly
content[].textstringRequired for type = textText prompt
content[].image_url.urlstringRequired for type = image_urlImage URL or uploaded asset ID
content[].video_url.urlstringRequired for type = video_urlVideo URL or uploaded asset ID. Video Base64 is not currently supported
content[].audio_url.urlstringRequired for type = audio_urlAudio URL or uploaded asset ID

Content constraints:

  • content must contain at least 1 item
  • Maximum 9 images
  • Maximum 3 videos
  • Maximum 3 audio files
  • text and audio can appear together only when at least one reference_image or reference_video is also present
  • Audio-only input is not supported
  • text + audio by itself is not supported
  • first_frame / last_frame cannot be mixed with reference_image / reference_video / reference_audio in the same request

Supported Seedance 2 input combinations:

  • text
  • optional text + image_url
  • optional text + video_url
  • optional text + image_url + audio_url
  • optional text + image_url + video_url
  • optional text + video_url + audio_url
  • optional text + image_url + video_url + audio_url

6. Legacy Fields No Longer Supported

The following fields now return 400:

  • input.prompt
  • input.image_url
  • input.image_urls
  • input.last_image_url
  • input.video_urls
  • input.audio_urls
  • input.negative_prompt
  • input.parameters
  • input.instances
  • input.service_tier
  • input.execution_expires_after
  • input.frames
  • input.camera_fixed
  • input.draft
  • input.content[].draft_task
  • Any undocumented custom fields

7. Request Examples

7.1 Text-to-Video

curl -X POST "https://direct.reachapi.ai/v1/vids/create" \
  -H "Authorization: Bearer YOUR_REACH_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "seedance-2",
    "input": {
      "content": [
        {
          "type": "text",
          "text": "A stylized travel film scene with dynamic camera movement"
        }
      ],
      "duration_seconds": 8,
      "resolution": "1080p",
      "aspect_ratio": "adaptive",
      "watermark": false
    }
  }'

7.2 Explicit Multimodal Reference Input

curl -X POST "https://direct.reachapi.ai/v1/vids/create" \
  -H "Authorization: Bearer YOUR_REACH_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "seedance-2",
    "input": {
      "content": [
        {
          "type": "text",
          "text": "A stylized travel film scene with dynamic camera movement"
        },
        {
          "type": "image_url",
          "role": "reference_image",
          "image_url": {
            "url": "https://cdn.example.com/scene-start.png"
          }
        },
        {
          "type": "video_url",
          "role": "reference_video",
          "video_url": {
            "url": "https://cdn.example.com/reference-motion.mp4"
          }
        }
      ],
      "duration_seconds": 8,
      "resolution": "1080p",
      "aspect_ratio": "adaptive",
      "watermark": false
    }
  }'

7.3 Reference Video Extension

curl -X POST "https://direct.reachapi.ai/v1/vids/create" \
  -H "Authorization: Bearer YOUR_REACH_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "seedance-2-fast",
    "input": {
      "content": [
        {
          "type": "text",
          "text": "Extend this clip into a faster action sequence"
        },
        {
          "type": "video_url",
          "role": "reference_video",
          "video_url": {
            "url": "https://cdn.example.com/reference-1.mp4"
          }
        }
      ],
      "duration_seconds": 6,
      "resolution": "480p",
      "aspect_ratio": "16:9",
      "extend_mode": "continue"
    }
  }'

8. Create Task Response

Example response:

{
  "code": 200,
  "msg": "",
  "status": "queued",
  "task_id": "task_xxx",
  "data": []
}

9. Query Task Status

Request example:

curl -X GET "https://direct.reachapi.ai/v1/tasks/task_xxx" \
  -H "Authorization: Bearer YOUR_REACH_API_KEY"

Successful response example:

{
  "code": 200,
  "msg": "",
  "status": "success",
  "task_id": "task_xxx",
  "data": [
    {
      "url": "https://cdn.example.com/generated/video-1.mp4"
    }
  ],
  "usage": {
    "input_tokens": 0,
    "output_tokens": 102960,
    "total_tokens": 102960
  },
  "cost": {
    "spend": 0.12
  }
}

Notes:

  • data contains video result items
  • ReachAPI guarantees data[].url for successful video outputs

10. Callback Delivery

If callback_url is provided when the task is created, the gateway sends a POST request to that URL when the task reaches a terminal state. The callback body uses the same JSON structure as the task query response.

11. Constraints and Notes

  • input.content is required
  • resolution is required
  • duration_seconds must be -1 or within 4-15
  • seedance-2-fast does not currently support 1080p
  • aspect_ratio supports adaptive
  • Input asset values must be non-empty and server-accessible when URLs are used

Seedance 1.5 API

Previous Page

Seedance 2.5 API

Next Page

On this page

1. API Overview
2. Authentication and Headers
3. Model Capability Summary
Billing and Usage
4. Create Task
4.1 Top-Level Parameters
4.2 input Parameters
5. input.content Structure
6. Legacy Fields No Longer Supported
7. Request Examples
7.1 Text-to-Video
7.2 Explicit Multimodal Reference Input
7.3 Reference Video Extension
8. Create Task Response
9. Query Task Status
10. Callback Delivery
11. Constraints and Notes