ChatGPT Free Users Can Now Generate Two AI Images Per Day on DALL-E

Free users of ChatGPT can now generate up to two images per day with DALL-E 3 after the AI image generator was previously exclusive to ChatGPT Plus subscribers.

Since its launch, DALL-E 3 has only been available to customers who pay for ChatGPT. Free users get access to ChatGPT 3.5 while subscribers get GPT 4 which is the more up-to-date large-language model.

It meant free users could not use DALL-E at all; not even the previous model, DALL-E 2, which was discontinued a few months ago. But now free users get two DALL-E 3 images per day.

Paying ChatGPT customers also have a limit of 50 images per day. However, two images every 24 hours might be appropriate for some users.

What is DALL-E 3 Like?

DALL-E 3 differs from its competitors such as Midjourney in that it is difficult to produce photorealistic images from it. Although some people might not be able to tell that it was generated by artificial intelligence, almost every output from DALL-E looks like an illustration or drawing.

That is a purposeful policy from OpenAI which says it wants to mitigate “misinformation and propaganda”. It is impossible to ask for a picture “in the style of” any given artist. Users who attempt this are greeted with “I was unable to generate the image you requested because it did not align with our content policy.”

However, as PetaPixel discovered back in March, it is still possible to recreate iconic artworks and photographs by writing a prompt that describes a famous image — rather than using the photographer’s name directly.

Photographers Can Opt Out of DALL-E

AI image generators such as DALL-E have caused great controversy since being launched because of the way the models are made. To create convincing artificial imagery, AI image generators must be trained on hundreds of millions of real-world images. Many of the owners of these images did not expressly consent to such usage.

OpenAI, in a possible bid to avoid future lawsuits, appears to have taken this into consideration by allowing creators to opt out of having their images used to train future models.

The removal form is available on its website. “We understand that some content owners may not want their publicly available works used to help teach our models,” reads the form.

Image credits: Header photo licensed via Depositphotos.