After an explosive launch, a viral trend, and some melted GPUs, the new image generation feature for ChatGPT is now available to free users. The feature originally launched on March 25 but because paid subscribers utterly flooded OpenAI with requests for Ghiblified images, CEO Sam Altman announced the next day that the rollout to free users would be delayed “a while.”
Luckily, it appears this delay is over just five days later — Altman has already published another X post saying that “image gen [is] now rolled out to all free users!”
I tried it out myself and although you do need to be logged in, the chatbot itself confirmed that it can now generate images. I always remember one random prompt that older versions of Dall-E would get confused about so I tried it out with this new version. Unfortunately, it still appears that everyone naturally plays their handheld gaming devices back to front in the AI universe.
However, the good news is ChatGPT was able to fix the image successfully after I pointed out the problem. While the otter appears to be having slightly less fun this time, it is still the same otter, so I’m calling it a win. Of course, there is the slight problem that this isn’t a Switch 2 — it’s an original Switch with “Switch 2” written on it — but we can forgive that. Plus, it gave the model a chance to show off its nice new text generation abilities.
The generation takes quite a while, it took a few minutes each for these two images, but it’s hard to say if that’s the norm or if it’s being slowed down by high activity. All of the hype around the new image generation feature has led to another load of new users for ChatGPT — Altman even said they got one million new users in one hour yesterday.
Altman also mentioned a few days ago that the free tier will be limited to three images per day and I can confirm that I got cut off after three generations. These limits might be loosened in a few weeks once all the excitement has died down or OpenAI finds a way to make the process a little more efficient.