OpenAI launches GPT-5.6 Sol Ultrafast mode

OpenAI has launched Ultrafast mode for its GPT-5.6 Sol model, now in limited preview through the OpenAI API. The service runs up to 14 times faster than standard processing and generates up to 750 output tokens per second, powered by Cerebras as part of their partnership on ultra-low-latency inference.

During the preview, customers are testing Ultrafast across coding, commerce, financial research, support, and other interactive applications in production. OpenAI will use the programme to assess where an order-of-magnitude increase in speed delivers the most value and how products change when models respond as quickly as users interact with them.

Internally, OpenAI is using Ultrafast for incident response, including reading logs, analysing traces, synthesising conversations, identifying follow-up checks, and helping prepare or validate fixes. The company is also applying it to research workflows involving knowledge searches, data queries, connected tools, and information synthesis, allowing research teams to complete multiple iterations during the workday instead of overnight.