OpenAI’s GPT-5.6 Sol on Ultrafast mode is available in limited preview to a select group of customers, launching first through the OpenAI API. The company says the service runs up to 14 times faster than Standard processing and generates up to 750 output tokens per second. Ultrafast is powered by Cerebras as part of the companies’ partnership on ultra-low-latency inference.
GPT-5.6 Sol Ultrafast and standard build a working 3D warehouse simulator from the same text prompt, side by side. (Source: OpenAI)
During the preview, customers are testing Ultrafast across coding, commerce, financial research, support, and other interactive applications in production environments. OpenAI will use the program to assess where an order-of-magnitude increase in speed delivers the most value and how products change when models can respond as quickly as users interact with them.
“The increase in speed brought by Cerebras is impressive. It enables different ways of using the models, and makes it practical for developers to work in a more focused and productive way alongside them,” John Crepezzi, AI Assistants, Jane Street, said.
Internally, OpenAI is using Ultrafast for incident response, including reading logs, analyzing traces, synthesizing conversations, identifying follow-up checks, and helping prepare or validate fixes. Faster inference shortens the time between observing a signal, testing a hypothesis, and selecting the next action, while engineers remain responsible for judgment and deployment.
The company is also applying Ultrafast to research workflows involving knowledge searches, data queries, connected tools, and information synthesis. OpenAI said research teams that might normally launch experiments overnight and review the results the following morning could instead complete multiple iterations during the workday.