← All models
Generate video from text and videos using NVIDIA's 2B Cosmos Post-Trained Model
Example
Details
- Accepts
- text + image + video
Pricing
- Price
- 25 cr / video
Prices in credits (1 credit = $0.01).
Data schema
Input
| Field | Type | Description |
|---|---|---|
| seed | — | Random seed for reproducible generation. |
| promptrequired | string | The text prompt describing the video to generate. |
| sync_mode | boolean | If `True`, the media will be returned as a data URI and the output data won't be available in the request history. |
| video_url | string | URL of the input video to use as conditioning. |
| num_frames | integer | Number of frames to generate. Must be between 9 and 93. |
| video_quality | stringenum: low, medium, high, maximum | The quality of the output video. |
| guidance_scale | number | Classifier-free guidance scale. Higher values increase prompt adherence. |
| negative_prompt | string | A negative prompt to guide generation away from undesired content. |
| video_output_type | stringenum: X264 (.mp4), VP9 (.webm), PRORES4444 (.mov), GIF (.gif) | The format of the output video. |
| num_inference_steps | integer | Number of denoising steps. More steps yield higher quality but take longer. |
| image_url | string | URL of the input image to use as first frame. |
Output
| Field | Type | Description |
|---|---|---|
| seed | integer | The random seed used for generation. |
| video | — | The generated video file. |