Cerebras launches ultrafast mode for OpenAI’s GPT-5.6 Sol

Cerebras Systems ↘ 11.85% announced Thursday that it is powering a new "Ultrafast" service tier in the OpenAI API for GPT-5.6 Sol, delivering up to 750 output tokens per second — up to 14 times faster than Standard processing with no loss in model intelligence.quiverquant+1

The launch, confirmed simultaneously by OpenAI and Cerebras, marks a deepening of the partnership between the two companies and represents a new frontier in the tradeoff between model capability and inference speed.

Speed Without Compromise

Available initially in limited preview to a select group of OpenAI customers, Ultrafast runs GPT-5.6 Sol at speeds that Cerebras claims are 5 times faster than Claude Opus 4.8 in Fast mode and 11 times faster than Claude Fable 5, based on output speeds reported by Artificial Analysis.investing+1

"GPT-5.6 Sol on Ultrafast is proof that speed and intelligence are no longer mutually exclusive," said Andrew Feldman, CEO and co-founder of Cerebras.quiverquant

Sachin Katti, VP Compute Strategy and GPT-Infra at OpenAI, said the companies are "exploring what becomes possible when customers can get the intelligence of our most capable models with significantly lower latency," adding that OpenAI plans to start with a small group of customers before expanding the service.openai+1

Benchmark Results

On Humanity's Last Exam, a 2,500-question benchmark spanning graduate-level subjects, GPT-5.6 Sol Ultrafast answered the full question set in 11 hours and 11 minutes — compared to more than three days for Claude Fable 5, achieving comparable accuracy nearly seven times faster. On GDP-Val, a benchmark of economically valuable knowledge-work tasks, Ultrafast delivered a 5.6 times end-to-end speedup with no quality degradation.cerebras+1

Architecture and Market Context

The speed advantage comes from Cerebras' Wafer-Scale Engine architecture, which keeps model weights on-chip with 44 GB of SRAM on each wafer-sized chip, eliminating the memory-bandwidth bottleneck that constrains GPU-based inference.cerebras

The announcement comes as Cerebras shares fell sharply Thursday despite the company raising its forecasts on AI demand following strong second-quarter growth. Multiple analysts raised their price targets on the stock, with UBS maintaining a Buy rating and Morgan Stanley maintaining an Overweight rating.marketscreener