Ever wished ChatGPT could pick up the pace a bit? OpenAI might have just heard you. The lab has introduced a new mode called Ultrafast, designed to dramatically speed up how its newest flagship model, GPT 5. 6 Sol, processes tasks. According to the company, Ultrafast operates at 14 times the speed of standard processing, delivering up to 750 output tokens per second—tokens being the individual chunks of text the model generates during conversations. In a blog post on Thursday, OpenAI explained, "Until now, getting real-time speed typically meant choosing a smaller or more specialized model. Ultrafast points to progress in a new direction: more useful work per second."

Rivals like Anthropic have also rolled out faster versions of their models. Claude offers a fast mode, though it doesn't quite match the velocity OpenAI is claiming here. OpenAI sees this high-speed iteration of GPT 5. 6 Sol as a fit for various corporate use cases, including incident response, customer service and support, financial market analysis, and e-commerce, among others. Ultrafast is currently in preview and is powered by OpenAI's collaboration with chipmaker Cerebras. For now, access is limited to a select group of customers, but OpenAI says it plans to broaden availability as "capacity grows."