Cerebras Powers New Ultrafast Mode for OpenAI GPT-5.6 Sol Models

Key Facts

In a move highlighting the intense competition for AI infrastructure, Cerebras has announced it is powering the new Ultrafast mode for OpenAI's GPT-5.6 Sol model API. This partnership aims to leverage Cerebras' specialized hardware to drastically reduce latency and increase processing speeds, introducing a high-speed service tier for frontier AI applications.

According to reports, this collaboration enables the model to run at speeds of up to 750 output tokens per second, which is approximately 14 times faster than standard processing while maintaining comparable accuracy. This development bolsters Cerebras' position as a cloud infrastructure provider, following data showing its core cloud revenue grew by 287% year-over-year, supported by expanding data center capacity.

Looking ahead, the partnership with OpenAI is viewed as a significant growth catalyst that could offset previous revenue misses and validate Cerebras' technology in a competitive market. While current price levels for CBRS are unavailable at this time, investors are monitoring how the planned 10x scaling of manufacturing capacity in 2026 will impact the company's long-term profitability.