Seedance 2.0 API
Generate cinematic AI video programmatically. Access Seedance 2.0 (SD 2.0) through Topview's REST API — text-to-video, image-to-video, and omni reference with up to 9 images, 3 videos, 3 audio files, and native audio sync.
Free tier available · No GPU required · Native audio included
const submit = await fetch('https://api.topview.ai/v1/common_task/omni_reference/task/submit', {
method: 'POST',
headers: {
Authorization: 'Bearer YOUR_API_KEY',
'Topview-Uid': 'YOUR_USER_ID',
'Content-Type': 'application/json'
},
body: JSON.stringify({
model: 'Standard',
prompt: 'Apply the style from <<<Image1>>> to the motion in <<<Video1>>>, cinematic lighting',
inputImages: [{ fileId: 'YOUR_IMAGE_FILE_ID', name: 'Image1' }],
inputVideos: [{ fileId: 'YOUR_VIDEO_FILE_ID', name: 'Video1' }],
aspectRatio: '16:9',
resolution: 720,
duration: 4,
generatingCount: 1
})
});
const { result } = await submit.json();
console.log(result.taskId);API Capabilities
Three endpoints cover every Seedance 2.0 workflow.
Text-to-Video
POST /v1/common_task/text2video/task/submit — Generate 480p or 720p video up to 15 seconds from a text prompt with native audio.
Image-to-Video V2
POST /v2/common_task/image2video/task/submit — Animate first frame + last frame images into video. Supports 480p/720p, 4–15s duration.
Omni Reference
POST /v1/common_task/omni_reference/task/submit — Mix up to 9 images + 3 videos + 3 audio files (12 total). Use the triple-bracket syntax to reference inputs in prompts.
Async Polling
All endpoints return a taskId. Poll the corresponding /task/query endpoint every 3–5 seconds to get status and video URL when complete.
Seedance 2.0 Highlights for Creators
Audio + Video, Generated Together
Create a complete clip in one pass so visuals and sound land on the same beat with less cleanup in post. This is one of Seedance 2.0's biggest workflow advantages over prompt-only competitors.
One-pass | Synced beats | Less post-fix
Multi-Shot in One Generate
Type the scene once and generate a structured sequence with cuts, angles, and pacing for up to 15 seconds in one pass. It is built for short narrative arcs, not just one isolated shot.
Up to 15s | Fewer stitches | Smoother arcs
Multi-Character Dialogue and Interaction
Stage scenes where multiple characters speak and react naturally, making conversations feel more watchable and less synthetic. This helps Seedance 2.0 work for ads, explainers, and story-first creator content.
Natural speech | Reactions | Watchable
Multi-Style Output
Switch quickly from cinematic drama to anime or ad-clean visuals without rebuilding your prompting workflow from scratch. Teams can test more concepts inside one production system.
Cinematic cuts | Guided pacing | Cohesive flow
Model Parameters
These are the core Seedance 2.0 specs most relevant to creators evaluating output quality, control depth, and production fit.
Seedance 2.0
Dual Branch Diffusion Transformer
~2 minutes
About 30% faster than the previous version
Max 12 files
Combined across all modalities
480p / 720p / 1080p
Flexible output options for faster drafts or higher-detail delivery
4s - 15s per shot
Extendable via multi-shot chaining
24fps
Cinema-standard output
16:9, 9:16, 1:1, 4:3, 3:4, 21:9
6 supported formats
Up to 9
Style, character, and scene references
Up to 3 clips
Motion transfer and camera reference
Up to 3 files
Beat sync, lip-sync, and atmosphere
Natural language
Detailed scene, pacing, and direction control
Dialogue + SFX + Music + Lip-sync
8+ languages, auto-generated
API Pricing
Seedance 2.0 (Standard) costs 1.2 credits/second at 720p and 0.6 credits/second at 480p. Sound is included at no extra cost. generatingCount multiplies the total.
View Full Pricing →Start Building with Seedance 2.0 API
Get your API key and start generating cinematic AI video with native audio in minutes.
Free tier available · REST API · Async polling