When I generate an image, in any browser, even incognito mode, after a single generation I’m stuck with this white button with a black square in it, and I can’t do anything to move forward. I’m going nuts, it’s halted my work flow - AGH! Any idea why, or how to fix this? It’s affecting my ability to customize the prompts. I spent time building recognition of a style and when I move to a new conversation it’s completely different. This is effecting both character continuity and style.
4 Likes
I’m having the same issue. To get around it I have to refresh the page after the image generation response and then submit my next prompt. Sometimes it already has my next prompt accepted and I have to click regenerate.
Also having the same issue on the mobile app.
Has any fix been created or just the page refresh as a work around?
After generating an image in ChatGPT, the interface gets stuck showing a white square (pause icon) instead of reverting to the usual arrow prompt. This blocks further input until I manually reload the page. I’ve tested this across multiple devices and browsers. It appears several users are experiencing the same issue based on posts in the OpenAI forums.
It isn’t working for me yet. Also page refresh does not always work.
I have the same issue, on two different operating systems (Linux and Windows) and also on my Android App. Another observation is that usually, at least for me, the first image generation goes normal. It is only upon the second image generation that I no longer get my input button back. On o4-mini, which is painfully slow for image generation, the ctrl-R trick seems to work. On the default 4o, there is nothing I can do to make it accept input again.
EDIT: One step further. I found if I ask for it to write a prompt first, “try that” next, then the ctrl-R trick works to bring back the input. It doesn’t otherwise. I do hope they fix this. It’s a terrible bug to just leave there for so long.
I also noticed that it ignores user preferences unless you remind it. So this is an extensive bug that suddenly has appeared.
Read my comment. I had the same problem, but did some experimenting and found a way around that. It sucks, but it seems to work. Basically, prompt it to show you the prompt for your image first. Then tell it “try that” You will still have the STOP button, but ctrl-R works after. At least for me. If it does, or doesn’t work, follow up so the devs may read it.
I’ve been having the same problem for a couple of days now. Refreshing the page allows me to keep going, but it gets stuck after every single image I generate. Also, response time for text has gotten significantly slower. Very unfortunate.
It works the first time, but not in subsequent generations.
I have the same issue, super annoying, refresh works but this is not how it suppose to work