I purchased GPTplus in order to use Dalle-3 and it worked at first however now when I ask it to create and image it say it is only a text based model and cannot create images and instead just lists a long description.
additionally when I refresh the page about 60% of the time it says I’m still on the free service and can only use gpt 3.5
the services I paid for aren’t working is this a known error and is it going to be resolved?
I am in the same situation. It keeps saying it is not capable of image generation.
I have tried to go the Explore Menu, and select DALLE, but the prompts are not send.
It seems that the Explore Menu doesn’t lead anywhere - I just tried the ChatGPT Classic and Sous Chef options with trivial prompts from there and these ones also froze. PHQ
Same, got the ChatGPT Plus update yesterday, created a couple of images and then it stopped working. Come on Open AI!
Yes, the same thing happened to me. I created the last image using the GPT-4 app just before the Open Dev Day, and after a while, there was, if I’m not mistaken, a brief update, which I think happened automatically (ui changed a bit). Since then, the output that GPT provides me is exactly what you mentioned about the inability to create images.
However, it’s worth noting that it works perfectly from the web. 
DELL-E is unavailable. When I type any prompt it always says: “DALL·E is currently experiencing high demand, so I wasn’t able to generate the image for you right now.” It is always the same answer for ANY prompt. I tried different accounts (plus subscription). And different browsers. And always the same response. Is there anyone facing the same issue?
I had one GPT that happened to be my first choice to check, and it had dalle unchecked as why it failed. Others did not and are working.
You will need to use a GPT, such as OpenAI’s “DALL-E”, also working.
If instead you are discussing the new image maker, it is not DALL-E any more. It’s just “ChatGPT makes images, without a good name” - but you must pick gpt-4o. And it is operational also, as well as the status page transformed by it.
OpenAI can also be arbitrary or selective with who gets through, or even use existing “shut-off” methods that don’t depict the actual maintenance or repair.
