Image Playground

The Image Playground is a dedicated space for testing and experimenting with GPUStack’s image generation APIs. It allows users to interactively explore the capabilities of different models, customize parameters, and review code examples for seamless API integration.

Generate Image

Prompt

You can input or randomly generate a prompt, then click the Submit button to generate an image.

Edit Image

Upload an image and highlight the areas you want to modify by painting over them. Then, enter a prompt and Submit. If no areas are painted, the entire image will be modified.

Save Mask

Click Save Mask to save the painted areas as a separate image.

Download Image

Click Download Image to save the edited image.

Select Model

You can select available models in GPUStack by clicking the model dropdown at the top-right corner of the playground UI.

Customize Parameters

You can customize the image generation parameters by switching between two API styles:

1. OpenAI Compatible mode
2. Advanced mode

Advanced Parameters

Parameter	Default	Description
Size	512x512	The size of the generated image in 'widthxheight' format.
Sample Method	euler_a	The sampler algorithm for image generation. Options include 'euler_a', 'euler', 'heun', 'dpm2', 'dpm++2s_a', 'dpm++2m', 'dpm++2mv2', 'ipndm', 'ipndm_v', and 'lcm'.
Schedule Method	discrete	The noise scheduling method.
Sampling Steps	10	The number of sampling steps to perform. Higher values may improve image quality at the cost of longer processing time.
Guidance	3.5	The scale for classifier-free guidance. A higher value increases adherence to the prompt.
CFG Scale	4.5	The scale for classifier-free guidance. A higher value increases adherence to the prompt.
Negative Prompt	(empty)	A negative prompt to specify what the image should avoid.
Seed	(empty)	Random seed.

View Code

After experimenting with prompts and parameters, click the View Code button to see how to call the API with the same inputs. Code examples are provided in curl, Python, and Node.js.