I was joking around about turning DALLE3 art into a 3d model, but it’s already kinda possible?
At least it is possible to create animations, all by itself.
I could fine-tune it, but decided to keep it as close to the original as possible.
See this little thread of mine;
Nicely done.
I used DALLE2 for sprites, but I haven’t tried with DALLE3 yet…
Yeah, I originally tried to get all frames in one image. I wonder if I could do separate images now with DALLE3… hrm… so many projects, so little time! 
You need an AI to help with all that 
(this is a complete sentence
What is the maximum number of simultaneous generations that can be performed via API?
Your rate-limit depends on your billing tier … which you can find here…
Anywhere from
dall-e-3 1 img/min
to
dall-e-3 15 img/min
They’ll likely be increased as well as things stabilize. Hope this helps.
You might not see your DALL-E 3 rate on the API limit page right now…although the model works. Get the headers with your $0.04-$0.12 picture.
The more severe dall-e-3 API limitation is numbers like 100 per day, like gpt-4-vision.
That’s placing you clearly in the “this is not a developer product to resell or implement” category, like tts-1-hd at 3 per minute.
DALL-E 2, I’m at 6000/hour, topping $100/hr with a bit of queueing to level it out. I can be labs.openai.com that works if I want. n=10 if you want ten variations on one prompt. And I’m nobody special.
Is it then the computation - or the content and partner exclusivity?
Wow! How about some of the posters with embossed emblems added? “AYOPA”!!
