Home / Tools / AI Image Prompt Generator

AI Image Prompt Generator

A free AI image prompt generator: describe your subject, tap preset styles, composition, lighting and colors, and get a prompt ready for Midjourney, Stable Diffusion, ComfyUI, Flux, ChatGPT or Gemini, in English or Chinese.

Everything happens in your browser. Nothing you type is uploaded.

Quick templates

Tap one to fill everything in, then change the text to your own idea.

1. Describe what you want

2. Tap the presets

Style
Shot and composition
Lighting
Color and mood · pick any
Quality boosters · pick any
Aspect ratio

How to use it

  1. Describe what you want in "Subject". The more specific, the better. Not sure where to start? Tap a quick template.
  2. Tap a style, shot, lighting and colors. All optional; tap again to clear.
  3. Choose where you will paste it and the prompt language, press Copy, and paste it into your image tool.

Tips for better prompts

  • Write the subject as who or what, what it looks like, and what it is doing: "a little girl in a red raincoat jumping in a puddle" beats "a girl".
  • Pick one style at a time. "Watercolor" plus "photorealistic" pull against each other and often look like neither.
  • Put unwanted things in "Things to leave out", not in the subject. Writing "no people" in the subject often adds people.
  • Generate several images from the same prompt and choose the best one. Every run is different.

Common questions

Should my prompt be in English or Chinese?

Midjourney, Stable Diffusion and ComfyUI work best in English. ChatGPT, Gemini and Chinese tools such as Jimeng also understand Chinese.

What is a negative prompt?

Text that tells the AI what not to draw, like blur, watermarks or extra fingers. Stable Diffusion and ComfyUI have a box for it, Midjourney uses --no, and for other tools it goes at the end of the prompt as "Avoid".

Is what I type saved anywhere?

No. The prompt is put together in your browser and never sent to a server.