Question 1

What language will the prompt use?

Accepted Answer

The generated prompt follows the current page language. If the page is in English, the result is English; if the page is in Simplified Chinese, the result is Chinese.

Question 2

What is the difference between the three AI modes?

Accepted Answer

Copy Mode writes a detailed prompt to recreate the source image. Style Extraction focuses on reusable visual style. General Description creates a clear and direct prompt describing the image content.

Question 3

What backend model should I configure?

Accepted Answer

Set the Image to Prompt model in the admin tool config. Use a NewAPI/OpenAI-compatible model with vision support, such as the newly added gpt-5.5.

Question 4

Can I edit the generated result?

Accepted Answer

Yes. The generated prompt is shown in an editable text area, so you can refine it before copying.

Generated result

FAQ

Ready to generate prompts from images?