Skip to main content

Edit with prompt tools not working

  • June 12, 2026
  • 2 replies
  • 35 views

JoyRipart

When I edit with prompt, it works in the preview but then it doesn’t apply to the image

 

June 14, 2026

Hi ​@JoyRipart — welcome to Figma Forum!

Thanks for flagging this, and for the follow-up noting that the issue is intermittent. I tested this on my end, including editing object colors within an image, and wasn't able to catch the same on-again-off-again behavior.

To help narrow things down, could you share a bit more:

  1. Prompt types. What kinds of prompts are you entering, and do some work more consistently than others?
  2. Platform and version. Is this happening in the browser, the Figma Desktop App, or both? If you're on the Desktop App, can you confirm you're running version 126.5.6 or later?
  3. Error details. When the tool doesn't respond, are you seeing any error messages — or does the image simply not react to the input?
  4. Model selection. Which model are you using (e.g. Nano Banana or ChatGPT Image)? Does switching to a different one make a difference?
“Select model” in lower left corner or prompt editor.

This will help me try to reproduce the issue and loop in the right team. If we need to go further, we can also take a look at the file directly.

I'll keep an eye out for your reply.

2 replies

JoyRipart
  • Author
  • New Member
  • June 12, 2026

it randomly started to work again sometimes and sometimes not


Tom Reem
Figmate
  • Figmate
  • June 14, 2026

Hi ​@JoyRipart — welcome to Figma Forum!

Thanks for flagging this, and for the follow-up noting that the issue is intermittent. I tested this on my end, including editing object colors within an image, and wasn't able to catch the same on-again-off-again behavior.

To help narrow things down, could you share a bit more:

  1. Prompt types. What kinds of prompts are you entering, and do some work more consistently than others?
  2. Platform and version. Is this happening in the browser, the Figma Desktop App, or both? If you're on the Desktop App, can you confirm you're running version 126.5.6 or later?
  3. Error details. When the tool doesn't respond, are you seeing any error messages — or does the image simply not react to the input?
  4. Model selection. Which model are you using (e.g. Nano Banana or ChatGPT Image)? Does switching to a different one make a difference?
“Select model” in lower left corner or prompt editor.

This will help me try to reproduce the issue and loop in the right team. If we need to go further, we can also take a look at the file directly.

I'll keep an eye out for your reply.