Skip to content

Local UI to run and train LLMs and diffusion models. Supports GGUF, MLX, Qwen3.8, DeepSeek-V4.1, MiniMax-H3, Gemma 4, FLUX and more.

Use model-neutral image generation tooltips (#10782)

Michael Han2fde5ab81dc0
diff --git a/studio/frontend/src/features/images/images-page.tsx b/studio/frontend/src/features/images/images-page.tsx
index 565e1561cbd..953d7eda76c 100644
--- a/studio/frontend/src/features/images/images-page.tsx
+++ b/studio/frontend/src/features/images/images-page.tsx
@@ -4155,7 +4155,7 @@ export function ImagesPage({
<div className="pt-2">
<SliderField
label="Steps"
- hint="9 is the recommended setting for Z-Image-Turbo. More steps rarely help."
+ hint="Number of denoising steps. Start with the selected model's default; more steps take longer and may not improve quality."
value={steps}
min={1}
max={50}
@@ -4165,7 +4165,7 @@ export function ImagesPage({
</div>
<SliderField
label="Guidance"
- hint="Keep this at 0 for Z-Image-Turbo. Higher values make its output worse. Other models use guidance."
+ hint="Controls how strongly the model follows the prompt. Start with the selected model's default; distilled models may require low or zero guidance."
value={guidance}
min={0}
max={15}