Z-Image-Turbo Workspace

Prompt Inputs

Bulk Mode

Prompts (One prompt per line)

Prompt Prefix

Prompt Suffix

Prompt History

No prompt history yet.

Model Select the active image generation model. Switching models will restart/reload the GPU resources.

Resolution Z-Image-Turbo is optimized for 1024x1024. Smaller dimensions run faster.

Inference Steps Z-Image-Turbo is distilled: use 9 steps. Higher values do not help quality.

Guidance Scale Must be 0.0. Higher values will corrupt or black out the image.

Seed (-1 for random) Set to -1 for random. Use a fixed integer to reproduce the composition.

0 Active Idle

# Prompt Status Time

🏖️

Queue is empty. Add prompts on the left to start generating!