The widespread enthusiasm surrounding ChatGPT’s enhanced and more readily accessible image generation features has compelled OpenAI to implement a “temporary” rate limit on image creation requests, as disclosed by CEO Sam Altman. “It’s incredibly gratifying to see such positive user engagement with images in ChatGPT, but our graphics processing units are under significant strain,” he conveyed in a recent post on the social media platform X.
Read: Bandwidth Blog & Smile 90.4FM Tech Tuesday: AI podcasts are coming!
While Altman refrained from specifying the precise parameters of the rate limit, he expressed optimism that this protective measure would be short-lived, as OpenAI endeavours to optimize its efficiency in managing the overwhelming influx of user requests.
The sheer volume of demand has already necessitated a delay in the rollout of the integrated image generation tool for users operating on ChatGPT’s complimentary tier. However, it appears that this pre-emptive action alone has proven insufficient to alleviate the infrastructural strain experienced by OpenAI. Altman indicated that free tier users will “shortly” be granted the ability to generate up to three images daily.
See also

OpenAI is harnessing the capabilities of its GPT-4o model to power this newly refined image generation utility. The outputs generated by this system often exhibit a heightened degree of realism, while simultaneously accommodating a broad spectrum of creative directions. Furthermore, the updated model has demonstrated marked improvements in addressing longstanding challenges, such as the accurate rendering of text within images.
OpenAI representatives characterized the advancements as a “substantial leap” beyond the capabilities of preceding models. Nevertheless, the imposition of rate limiting underscores the formidable computational resources, both in terms of processing power and energy consumption, that are prerequisites for image generation within the ChatGPT environment.
Crédito: Link de origem