OpenAI Accelerates GPT-5.6 Sol to be 14x Faster: 750 Tokens/s Coming to API First

According to TrendForce's Beating Monitor, OpenAI has started previewing the Ultrafast mode of GPT-5.6 Sol. It is powered by Cerebras' computing power, with a maximum output speed of 750 tokens/s, 14 times faster than the Standard mode. Currently, it is only available to a small number of API clients.

OpenAI has tested Ultrafast in scenarios such as troubleshooting, research, customer service, financial analysis, and Agent development. For Agents that need to call the model repeatedly, the accelerated single-response time can directly reduce the overall task execution time.

This is a step up from the Fast mode introduced by OpenAI at the end of July. The Fast mode is up to 2.5 times faster than the Standard mode, while Ultrafast pushes the limit to 14 times. OpenAI has not yet disclosed the pricing for Ultrafast, nor has it announced when it will enter ChatGPT.