Image Generate
Start generating images from a prompt.
One generation runs at a time. Poll GET /api/v1/images/{generation_id} for progress and finished images.
Returns 409 while another generation or a GPU job is running.
application/json- body
model_id*Model IdImage model, from GET /api/v1/images/models
length <= 64prompt*PromptWhat to draw
1 <= length <= 2000negative_prompt?|What to avoid. Only for models whose negative_prompt is true
count?|Number of images. Defaults to the model's images value
aspect?|Image shape. Defaults to the model's size
seed?|Seed for the first image. Each next image uses the next seed
steps?|Sampling steps, within the model's steps_range. Defaults to the model's steps
guidance?|Prompt guidance, within the model's guidance_range. Defaults to the model's guidance
Successful Response
response?unknownapplication/jsoncurl -X POST 'http://127.0.0.1:8100/api/v1/images' \ -H 'Content-Type: application/json' \ -d '{ "model_id": "string", "prompt": "string"}'nullImage Models GET
List the image models with whether each is installed and fits this computer. `images` is how many images a request makes when `count` is left out, and `size` is the default width and height. `steps_range` and `guidance_range` are the values a request may set, or null when the model's value is fixed. `download_gb` is what the first generation downloads.
Image Status GET
Get a generation's status, progress and finished images. `status` is running, done, failed or cancelled. Each image lists its seed and the URL of its PNG.