OpenAI introduces ‘Ultrafast,’ a new mode that makes GPT 5.6 Sol work at 14x the speed
OpenAI says Ultrafast can generate up to 750 output tokens per second in a limited preview.
The mode is built for GPT-5.6 Sol and is pitched as a way to get real-time speed without moving to a smaller model. OpenAI names incident response, customer support, financial analysis, and e-commerce as likely corporate uses. The preview is powered by its Cerebras partnership and is available only to a small customer group for now. Access is expected to widen as capacity grows. TechCrunch AI's note
The mode is built for GPT-5.6 Sol and is pitched as a way to get real-time speed without moving to a smaller model. OpenAI names incident response, customer support, financial analysis, and e-commerce as likely corporate uses. The preview is powered by its Cerebras partnership and is available only to a small customer group for now. Access is expected to widen as capacity grows. TechCrunch AI's note
score 8