POST
v1
/
replicate
/
imageupscale
curl --request POST \
  --url http://localhost:9000/api/v1/replicate/imageupscale \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: multipart/form-data' \
  --form 'prompt=<string>' \
  --form 'negative_prompt=<string>' \
  --form scale_factor=123 \
  --form dynamic=123 \
  --form creativity=123 \
  --form resemblance=123 \
  --form tiling_width=123 \
  --form tiling_height=123 \
  --form 'sd_model=<string>' \
  --form 'scheduler=<string>' \
  --form num_inference_steps=123 \
  --form downscaling=true \
  --form downscaling_resolution=123 \
  --form sharpen=123 \
  --form 'output_format=<string>'

​
Query

​
model
string
default:"philz1337x/clarity-upscaler"
required

The image model to be used;

​
Body

​
image
file
required

Input image

​
prompt
string

Prompt

​
negative_prompt
string

Negative Prompt

​
scale_factor
float
default:"2.0"

Scale factor

​
dynamic
float
default:"6.0"

HDR, try from 3 - 9

​
creativity
float
default:"0.35"

Creativity, try from 0.3 - 0.9

​
resemblance
float
default:"0.6"

Resemblance, try from 0.3 - 1.6

​
tiling_width
int
default:"112"

Fractality, set lower tile width for a high Fractality

​
tiling_height
int
default:"114"

Fractality, set lower tile height for a high Fractality

​
sd_model
string
default:"uggernaut_reborn.safetensors [338b85bc4f]"

Stable Diffusion model checkpoint

​
scheduler
string
default:"DPM++ 3M SDE Karras"

scheduler

​
num_inference_steps
int
default:"18"

Number of denoising steps

​
downscaling
boolean
default:"false"

Downscale the image before upscaling. Can improve quality and speed for images with high resolution but lower quality

​
downscaling_resolution
int
default:"768"

Downscaling resolution

​
sharpen
float
default:"0.0"

Sharpen the image after upscaling. The higher the value, the more sharpening is applied. 0 for no sharpening

​
output_format
string
default:"png"

Format of the output images