Seedance v1.0 Lite Image-to-Video
Seedance v1.0 Lite Image-to-Video animates a still image into video. The source photograph anchors visual identity while a text prompt directs motion, camera work, and scene evolution. It's a cost-optimized path from static asset to moving content.
import { experimental_generateVideo as generateVideo } from 'ai';
const result = await generateVideo({ model: 'bytedance/seedance-v1.0-lite-i2v', prompt: 'A serene mountain lake at sunrise.'});Playground
Try out Seedance v1.0 Lite Image-to-Video by ByteDance. Usage is billed to your team at API rates. Free users (those who haven't made a payment) get $5 of credits every 30 days.
Qwen 3 32B
Ask Qwen 3 32B anything to try it out.
Providers
Route requests across multiple providers. Copy a provider slug to set your preference. Visit the docs for more info. Using a provider means you agree to their terms, listed under Legal.
| Provider |
|---|
 ByteDance | — | — | 06/01/2025 |
More models by ByteDance
| Model |
|---|
| — | — | | 04/14/2026 | ||||||||||
| — | — | | 04/14/2026 | ||||||||||
| — | — | | 10/24/2025 | ||||||||||
| 256K | 1.2s | 80tps | $0.25/M | $2.00/M | Read: $0.05/M Write: — | — | — | +2 | | 09/01/2025 | |||
| 256K | 1.1s | 97tps | $0.25/M | $2.00/M | Read: $0.05/M Write: — | — | — | +2 | | 09/01/2025 | |||
| — | — | | 06/11/2025 |
About Seedance v1.0 Lite Image-to-Video
Most creative teams already have the still image. The product shot is done, the portrait approved, the illustration finalized. Seedance v1.0 Lite Image-to-Video accepts that finished image alongside a text prompt and generates a video clip where the subject moves, the camera shifts, and the scene evolves while maintaining visual consistency with the original photograph.
This differs from text-to-video (T2V) generation. A T2V model generates every visual element from scratch based on text alone. With image-to-video (I2V), the source image anchors the output. Colors, textures, spatial layout, and subject appearance carry over from the photograph. The model adds plausible motion and temporal continuity on top of that foundation.
Seedance v1.0 Lite Image-to-Video shares the unified pre-training framework described in the Seedance 1.0 tech report. That architecture handles both T2V and I2V through binary task masks, so the motion modeling, structural stability, and multi-shot awareness that earned Seedance first-place rankings on Artificial Analysis leaderboards are present here. Output spans 720p and 1080p at 24 fps, with durations from two to 12 seconds and seven aspect ratios: 16:9, 9:16, 1:1, 4:3, 3:4, 21:9, and 9:21.
See https://console.byteplus.com/ark/region:ark+ap-southeast-1/model/detail?Id=seedance-1-0-lite-i2v for methodology and benchmark context.
The Lite tier targets volume. Teams animating entire product catalogs or generating dozens of social video variants from one approved photograph can do so at a lower per-clip cost than Pro-tier options.
What To Consider When Choosing a Provider
- Configuration: Before you scale an image-to-video pipeline to production, verify that your reference images meet the model's expected input format and resolution. Higher-resolution sources usually preserve identity better in the output. Compare rates (listed video cost N/A; tier lines N/A).
- Zero Data Retention: AI Gateway does not currently support Zero Data Retention for this model. See the documentation for models that support ZDR.
- Authentication: AI Gateway authenticates requests using an API key or OIDC token. You do not need to manage provider credentials directly.
When to Use Seedance v1.0 Lite Image-to-Video
Best For
- E-commerce product animation: Turn approved catalog photography into short video clips that show the product in motion
- Social media content: Convert an existing brand image or illustration into a short-form video for platform distribution
- Animation direction exploration: Test different prompts on a single source image to see how the same photograph can move in different ways
- Marketing asset pipelines: Generate video variants from a locked hero image without commissioning new shoots or renders
Consider Alternatives When
- No reference image: Use a T2V variant when the video should come entirely from a text description
- Cinematic output: Seedance 1.0 Pro provides a higher motion-fidelity ceiling with advanced directorial controls
- Video-based source: Reference-to-video models are designed for workflows where the source material is a video clip rather than a still image
Conclusion
Seedance v1.0 Lite Image-to-Video turns the still image you already have into the video you need. By anchoring on a source photograph rather than building from text alone, it preserves visual identity with less prompt engineering.