lithovas.
← All models
✦

Seedance 2

bytedance/seedance-2Commercial use

Seedance 2.0 Standard is a flagship video model. One model slug covers text-to-video, single-frame and first/last-frame I2V, and omni reference (image + video + audio). Available through the unified 97AI.PRO async API.

💰Pricing: 4K no video input 218.4 credits / s (≈$1.092), 4K with video input 134.4 credits / s (≈$0.672), 1080p with video input 65.1 credits / s (≈$0.3255), 1080p no video input 107.1 credits / s (≈$0.5355), 720p no video input 43.05 credits / s (≈$0.2153), 720p with video input 26.25 credits / s (≈$0.1313), 480p no video input 19.95 credits / s (≈$0.0998), 480p with video input 12.08 credits / s (≈$0.0599)
24H STATUS MONITORSuccess: 100%
No requests in the last 24h · Operational
Seedance 2 reference media limits▾
These values are public validation rules for the upstream interface currently connected to 97AI.PRO. They are not permanent official model limits. Available ranges can vary by route; use the API response as the final authority.

Reference images

Formats
JPEG, PNG, WEBP, BMP, TIFF, GIF
Width / height
300-6000 px
Aspect ratio
0.4-2.5
File size
< 30 MB

Reference videos

Formats
MP4, MOV
Resolution tiers
480p, 720p
Duration per file
2-15 s
Count / total duration
Up to 3 / no more than 15 s total
Width / height
300-6000 px
Total pixels
409600-927408
Aspect ratio
0.4-2.5
File size
<= 50 MB
Frame rate
24-60 FPS

Reference audio

Formats
WAV, MP3
Duration per file
2-15 s
Count / total duration
Up to 3 / no more than 15 s total
File size
<= 15 MB
Sample rate / channels
Not publicly disclosed
Audio codec
Not publicly disclosed

Output video

resolution
480p, 720p, 1080p, 4k
aspect_ratio
1:1, 4:3, 3:4, 16:9, 9:16, 21:9, adaptive
Exact pixel dimensions
Not publicly disclosed; a fixed width and height are not guaranteed
Input
MODEL
Seedance 2
bytedance/seedance-2

Text prompt (max 30000 characters). Use @image1, @video1, @audio1 when reference_* fields are set.

First-frame image URL or asset://{assetId}. Mutually exclusive with reference_* fields.

Last-frame image URL or asset://{assetId}. Requires first_frame_url.

Portrait-library route: ingest hint for first_frame_url. Default skip (direct URL). Use virtual/real to upload portraits to the asset library.

Portrait-library route: ingest hint for last_frame_url. Default skip (direct URL). Use virtual/real to upload portraits to the asset library.

Reference image URLs (max 9). Mutually exclusive with strict first/last frame mode.

Per-image ingest hint aligned with reference_image_urls: virtual (portrait library), real (alias), skip (direct URL; default when omitted).

Structured reference images: [{ url, portrait: "virtual"|"real"|"skip" }]. Alternative to reference_image_urls + reference_portrait_kinds.

Portrait URLs always uploaded to the virtual portrait library before generation.

Portrait URLs (including real-person photos) uploaded to the virtual portrait library before generation.

Reference video URLs (max 3). Never uploaded to the asset library.

Reference audio URLs (max 3). Never uploaded to the asset library.

Generate synchronized audio with the video.

Return the generated video's last frame.

Output video resolution (480p–4K).

Output aspect ratio. Use adaptive for single-frame or first/last-frame I2V.

Video duration from 4 to 15 seconds.

Video output format.

Enable online search for the generation.

Enable content safety checking (97AI relay; not forwarded to upstream output moderation).

Output
output type: video
Result will appear here.