Video Models
Video Models are the AI models available in the Video Node's generation panel.
Available models
Seedance 2.0
The default model. Supports all generation modes and the widest range of reference inputs.
Resolutions: 480p, 720p, 1080p
Input limits:
- Image Nodes: up to 9
- Video Nodes: up to 3
Supported modes: Create Video, First & Last Frame, Omni-Reference
Kling O3
Resolutions: 720p, 1080p
Input limits:
- Image Nodes: up to 7
- Video Nodes: up to 1
Supported modes: Create Video, First & Last Frame, Omni-Reference
If you connect more Image or Video Nodes than Kling O3 supports, the system automatically switches to Seedance 2.0.
Kling 3.0
Resolutions: 720p, 1080p
Input limits:
- Image Nodes: up to 7
- Video Nodes: not supported
Supported modes: Create Video, First & Last Frame, Omni-Reference
If you connect an 8th Image Node or any Video Node while Kling 3.0 is selected, the system automatically switches to a compatible model.
Wan 2.7
Resolutions: 720p, 1080p
Generates video with audio by default. Subscription required.
Additional models
| Model | Subscription |
|---|---|
| Google Veo 3.1 | Yes |
| Kling 2.1 Master | Yes |
| Kling 2.0 Master | Yes |
| Hailuo 02 | Yes |
| Luma Ray 2 | Yes |
| Seedance 1.0 Pro | Yes |
| Wan 2.5 | Yes |
| Wan v2.2 5b | No |
Generation modes and model compatibility
| Mode | Compatible models |
|---|---|
| Create Video | All models |
| First & Last Frame | Seedance 2.0, Kling O3, Kling 3.0 |
| Omni-Reference | Seedance 2.0, Kling O3, Kling 3.0 |
Tip: The Generate button shows the credit cost for the current model, resolution, and duration before you confirm.
Updated about 2 months ago