You're looking at a specific version of this model. Jump to the model overview.
Input schema
The fields you can use to run this model with an API. If you don’t give a value for a field its default value will be used.
| Field | Type | Default value | Description |
|---|---|---|---|
| image |
string
|
Input image to upscale
|
|
| user_prompt |
string
|
|
User prompt to guide the super-resolution
|
| positive_prompt |
string
|
clean, high-resolution, 8k, best quality, masterpiece
|
Positive prompt
|
| negative_prompt |
string
|
dotted, noise, blur, lowres, oversmooth, longbody, bad anatomy, bad hands, missing fingers, extra digit, fewer digits, cropped, worst quality, low quality
|
Negative prompt
|
| num_inference_steps |
integer
|
4
Min: 1 Max: 8 |
Number of inference steps (SD Turbo optimized for 1-4 steps on Replicate)
|
| scale_factor |
integer
|
4
Min: 1 Max: 6 |
Super-resolution scale factor (max 4x recommended for Replicate)
|
| cfg_scale |
number
|
1
Min: 0.5 Max: 1.5 |
Classifier Free Guidance Scale (SD Turbo optimized at 1.0)
|
| use_kds |
boolean
|
True
|
Use Kernel Density Steering
|
| bandwidth |
number
|
0.1
Min: 0.1 Max: 0.8 |
Bandwidth for KDS
|
| num_particles |
integer
|
10
Min: 1 Max: 16 |
Number of particles for KDS
|
| seed |
integer
|
231
|
Random seed for reproducibility
|
| latent_tiled_size |
integer
|
320
Min: 128 Max: 480 |
Diffusion tile size for memory management
|
| latent_tiled_overlap |
integer
|
4
Min: 4 Max: 16 |
Diffusion tile overlap
|
Output schema
The shape of the response you’ll get when you run this model with an API.
{'format': 'uri', 'title': 'Output', 'type': 'string'}