ChatGPT’s updated image generator is now free for all — but there’s a catch

OpenAI’s latest GPT‑4o image generation tool can now be accessed by all ChatGPT users, regardless of whether they have a subscription or not. However, there is a limit to how many images you can generate in a day

The new image generator was initially only available to ChatGPT Plus subscribers. However, in a recent X post, OpenAI CEO Sam Altman announced that the tool is now available to all users.

Altman has previously stated on X that users of ChatGPT’s free tier could generate up to three images a day before they have to stop, or upgrade to one of the paid subscriptions. However, when I tested it, I could only create one before running into complications. When it tried to create a second image, I was told that I needed to wait for the limit to reset before I could create another.

When it comes to image generation, there’s little doubt that ChatGPT’s GPT-4o image generation tool is one of the best. In our tests, we found seven great examples of just how detailed it could be when generating images.

While access to the feature is great, the issue with limits indicates that OpenAI is still looking for a solution to its recent issues regarding the app. While it wasn’t a universal issue, many stated that the fault was due to how many people were trying to generate images at once. As such, our issues with the limit numbers might be a holdover from that.

Increased access will increase the debate regarding what’s created

While access to more powerful image generation might be great for some, others might despair at the possibility of even more generated art that arguably steals from other creators. Recently, we’ve seen a wealth of AI-generated art that apes the style of Hayao Miyazaki, the animator and filmmaker behind Studio Ghibli.

This has left many fans upset, and has fed into the continued debate around AI generation, creator rights, and whether it is good or bad for artists.

We’ve seen similar arguments in the past regarding OpenAI, especially regarding how and what it uses to train its AI. For instance, last year, we saw Open AI drop its Sky voice after a legal confrontation with actor Scarlett Johansson, who claimed the voice was copied from hers without her permission.

As it stands, ChatGPT’s GPT-4o tool remains one of the best means for generating images with ease. However, its existence only adds fuel to the debate around AI-generated images while AI companies, creatives and users try to negotiate a consensus.

More from Tom's Guide

• iOS 19 software support rumor spells the end for three different iPhones — here's the models that will (and won't) get the update
• April Fools’ Day 2025 live — the best (and worst) jokes and pranks
• I used ChatGPT 4o's new AI image generator to turn my family into catoons — the results blew me away