OpenAI is pausing the broader rollout of its latest image capabilities in ChatGPT after encountering unexpectedly high demand, according to CEO Sam Altman. The announcement was made Wednesday via a post on X (formerly Twitter).

“Images in ChatGPT are wayyyy more popular than we expected (and we had pretty high expectations),” Altman wrote.

GPT-4o’s Image Tools Initially Reserved for Paid Users

The delay comes just one day after OpenAI introduced its new GPT-4o model, which includes native support for image generation and editing. The update allows users to upload, modify, and interact with images directly within ChatGPT.

In a blog post published Tuesday, OpenAI said these new features would roll out to all ChatGPT users “soon.” However, as of Wednesday, the capabilities remain limited to paying subscribers—specifically those on ChatGPT Plus, Pro, and Teams plans.

Ongoing Capacity Challenges at OpenAI

This isn’t the first time OpenAI has scaled back or delayed product rollouts due to infrastructure limitations. The company has repeatedly cited insufficient compute capacity as a bottleneck for launching new features.

Earlier this year, OpenAI faced a similar issue following the debut of its text-to-video model Sora. Signups for the product were temporarily paused just days after launch due to demand.

Looking ahead, OpenAI is betting big on infrastructure to address these recurring issues. The company is reportedly planning a $500 billion initiative, dubbed Stargate, to build one of the largest data centers in the world in partnership with Microsoft. The project is aimed at dramatically increasing the compute power available for AI workloads.

Image: OpenAI