OpenAI launches 14× faster GPT-5.6 Sol via Cerebras-powered Ultrafast tier

OpenAI has previewed its Ultrafast service tier for API users, enabling GPT-5.6 Sol to operate at speeds up to 14 times faster than standard implementations. The performance boost, achieved through partnership with Cerebras hardware, delivers up to 750 output tokens per second – a significant throughput increase for high-demand AI applications. This development primarily benefits enterprise users requiring low-latency responses for real-time applications. The technical achievement underscores continuing hardware-software co-design advancements in large language model deployment, though exact availability timelines and pricing remain unspecified.