The preview mode delivers up to 750 output tokens per second and runs on Cerebras hardware.
OpenAI has rolled out a new mode called Ultrafast for its flagship model, GPT-5.6 Sol. The mode runs at 14x the speed of standard processing and delivers up to 750 output tokens per second. OpenAI announced the preview on Thursday and positioned it as a step toward more useful work per second.
OpenAI’s partnership with chipmaker Cerebras powers Ultrafast. OpenAI currently limits the preview to a small group of customers. OpenAI’s competitors, such as Anthropic, have launched accelerated versions of their models, including a fast mode for Claude, though it does not match the speed that OpenAI offers.
For builders and operators, this speed enables new workflows in incident response, customer service, financial market analysis, and e-commerce. The token throughput supports real-time interactions without requiring a smaller or specialized model. OpenAI says Ultrafast represents progress in the direction of more useful work per second.
OpenAI will expand access to Ultrafast as capacity grows. The company currently limits the preview to a small group of customers, but it plans to widen availability. Builders should watch for further rollout details and real-world performance as the mode scales.
What matters
- OpenAI rolled out Ultrafast, a preview mode for GPT-5.6 Sol that runs at 14x standard speed.
- Builders can use the speed boost for incident response, customer service, and market analysis.
- OpenAI will expand access to Ultrafast as capacity grows beyond the initial preview.
Why it matters
OpenAI will expand access to Ultrafast as capacity grows beyond the initial preview.
This GenAI News article was prepared in original wording using reporting and materials published by TechCrunch AI. Source reference: https://techcrunch.com/2026/08/13/openai-introduces-ultrafast-a-new-mode-that-makes-gpt-5-6-sol-work-at-14x-the-speed/.
Drafted by the GenAI News review pipeline.