Seedance 2.0
Seedance 2.0 is ByteDance's multimodal AI video model on KIE, combining text prompts with optional image, video, and audio references for controllable cinematic video generation.
Seedance 2.0 Multimodal Video
The text prompt or description for the video.
Optional image URL used as the first frame.
Drag and drop an image, or click to browse
Supports PNG, JPEG, WebP. Max 30MB. You can also paste from clipboard (Ctrl+V).
Optional image URL used as the last frame.
Drag and drop an image, or click to browse
Supports PNG, JPEG, WebP. Max 30MB. You can also paste from clipboard (Ctrl+V).
A list of input image URLs. Up to 9 images.
Drag and drop an image, or click to browse
Supports PNG, JPEG, WebP. Max 30MB. You can also paste from clipboard (Ctrl+V).
A list of input video URLs. Up to 3 videos, 15 seconds total.
Drag and drop a video, or click to browse
Supports MP4, MOV, WebM. Max 50MB.
A list of input audio files or URLs. Up to 3 audio files, 15 seconds total.
Drag and drop audio, or click to browse
Supports MP3, WAV, X-WAV, AAC, MP4, OGG. Max 15MB.
Whether to generate AI audio synchronized with the video.
The output video resolution.
The aspect ratio of the generated video.
Video duration in seconds.
4–15
Result
Results will appear here
Cost: $0.205/second
Seedance 2.0 API Implementation Guide
Get API Key
Generate your API key from the dashboard and set it as an environment variable.
export MODELHUNTER_KEY="your_api_key"Make Your First Call
Follow the API documentation for complete code examples in cURL, JavaScript, and Python, including request format, polling, and error handling.
View API Documentation →Seedance 2.0 Key Info
Capabilities
- Multimodal Video
Pricing
Price per second by resolution
| Option | 480p | 720p | 1080p |
|---|---|---|---|
| no video reference | $0.095/second | $0.205/second | $0.51/second |
| with video reference, input + output seconds | $0.0575/second | $0.125/second | $0.31/second |
No-video-reference pricing: $0.095/s @480p, $0.205/s @720p, $0.51/s @1080p. Video-reference requests: $0.0575/s @480p, $0.125/s @720p, $0.31/s @1080p, billed on input + output seconds.
Use Cases
Output Formats
Seedance 2.0 API Documentation
Run the playground first, then open the full API docs for integration details.
About
Generate video from a prompt with optional first/last frame, image references, video references, and audio references.
Code Examples
1curl -X POST https://api.modelhunter.ai/api/v1/seedance/multimodal-video \2-H "Authorization: Bearer YOUR_API_KEY" \3-H "Content-Type: application/json" \4-d '{5"model": "bytedance/seedance-2",6"input": {7"prompt": "A futuristic cityscape at sunset with neon signs",8"generate_audio": "true",9"resolution": "720p"10}11}'
Schema
View Full API DocsInput
modelModel ID for this capability.
Value: bytedance/seedance-2
promptThe text prompt or description for the video.
first_frame_urlOptional image URL used as the first frame.
last_frame_urlOptional image URL used as the last frame.
reference_image_urlsA list of input image URLs. Up to 9 images.
reference_video_urlsA list of input video URLs. Up to 3 videos, 15 seconds total.
reference_audio_urlsA list of input audio files or URLs. Up to 3 audio files, 15 seconds total.
generate_audioWhether to generate AI audio synchronized with the video.
Default: true
resolutionThe output video resolution.
Options: 480p, 720p, 1080p
Default: 720p
aspect_ratioThe aspect ratio of the generated video.
Options: 16:9, 4:3, 1:1, 3:4, 9:16, 21:9
Default: 16:9
durationVideo duration in seconds.
Default: 15
Range: 4 to 15
web_searchUse online search.
Default: false
webhookUrlURL to receive a webhook notification when the task completes.
metadataCustom key-value metadata to attach to the task for tracking.
Response (HTTP 202)
successWhether the request was accepted successfully.
data.idUnique task ID. Use this to poll GET /api/v1/tasks/{id} for results.
data.statusTask status: pending, processing, completed, failed.
data.typeGeneration type: multimodal-video.
data.providerProvider name: seedance.
data.modelModel ID used for generation.
data.created_atISO 8601 timestamp of when the task was created.
data.estimated_secondsEstimated time in seconds until the task completes.