← All models
FlashVSR+ upscales video by up to 4x with a one-step diffusion model, offering a fast Turbo mode and a higher-quality Base mode. Output reaches about 4K, and the input audio track is preserved by default.
Details
- Accepts
- video
Pricing
- Price
- 0.063 cr / megapixel
Prices in credits (1 credit = $0.01).
Data schema
Input
| Field | Type | Description |
|---|---|---|
| seed | — | The random seed used for the generation process. |
| model | stringenum: turbo, base, v1 | turbo: fastest (hybrid sliding attention). base: highest quality. v1: FlashVSR v1.1 weights with the base configuration. |
| color_fix | boolean | Apply low-frequency wavelet color correction. |
| sync_mode | boolean | If `True`, the media will be returned inline and not stored in history. |
| video_urlrequired | string | The input video to upscale. |
| video_quality | stringenum: low, medium, high, maximum | Output encoding quality. |
| preserve_audio | boolean | Copy the input audio track into the output. |
| upscale_factor | number | Spatial upscaling factor. Output is capped at about 4K. |
| video_write_mode | stringenum: fast, balanced, small | Output write mode. |
| video_output_type | stringenum: X264 (.mp4), VP9 (.webm), PRORES4444 (.mov), GIF (.gif) | Output container/codec. |
| temporal_compression | integerenum: 1, 2, 4 | Input frames per latent frame. Use 2 or 1 for large motion; inference cost rises about 2x or 4x. |
| enable_safety_checker | boolean | Authorized callers may set false to request relaxed screening. |
Output
| Field | Type | Description |
|---|---|---|
| seed | integer | The random seed used for the generation process. |
| video | — | The upscaled video. |