Z.aiCogVideoX
CogVideoX1.5-5B
CogVideoX1.5 provides five-billion-parameter text-to-video and image-to-video checkpoints. Both generate five- or ten-second clips at 16 frames per second.
Restricted access
Practical use
Practical guidance. Select the checkpoint from the intended input: a written scene or an existing image. Check subject continuity throughout the clip and review the commercial terms before using the weights in a customer-facing service.
Specifications & access
| Detail | Value and conditions |
|---|---|
| Status | Restricted access |
| Access | Downloadable |
| Model identifiers | zai-org/CogVideoX1.5-5Bzai-org/CogVideoX1.5-5B-I2V |
| Resolution | Text-to-video: 1360 × 768. Image-to-video: the short side is 768 and the long side can reach 1360.CogVideoX1.5-5B model card |
| Licence | Commercial use requires registration and a basic commercial licence. More than one million monthly service visits requires additional permission.CogVideoX licence |
Related models
- CogVideoX-3
Z.ai · CogVideoX
A hosted video model with text, image and endpoint-frame inputs.
Launch date not confirmedAPI