So I’ve been thinking to get api key for a product which I’m building. But I wanna know the mechanism how it works. Please bear with me I’m a newbie developer and want to try out things. 
Hey I want to know same thing
Can we do collaboration
If so let’s do some texting
The API is not ChatGPT, they are different products.
The current all-new image gen on ChatGPT has not yet been released on the API.
That said DALL-E is accessible from the API currently and the prices are here:
https://platform.openai.com/docs/pricing#image-generation
I imagine that if & when the new image generator is released on the API (here’s hoping!), the price will be added there.
1 Like
What I’d like to know is whether the pricing will be based on tokens, similar to text, and if it will be more affordable than DALL·E 3 via the API. Given the longer render times, I suspect the cost will be significantly higher, which could make it unaffordable—despite its potential as a powerful tool for automated API call sequences. Unless, of course, they introduce something like a “turbo” image model to offset that.
2 Likes
I suspect you are right. It will be much more expensive. I think that’s clear from how long they take to render. It’s not like OpenAI are deliberately slowing it down.
Then again, it is also much better and more useful. So it will be worth more expenditure.
Yeah but 3$ for 100 images on dalle-3 for someone who is used to rendering 3000 images a day locally is already nuts, that is 2700$ a month and then I still would need to run them all through flux + upscaling. If it’s 3-10x the price then this image generator is too stupidly priced to even think about using.
So is it not possible right now to post image to 4o model and get image in result?
I’ve seen some people using 4o to generate those Studio Ghibli style images
Not possible on the API yet I believe.
Try it 
Yeah that is regular user gpt not API gpt.