
OpenAI has previewed Ultrafast, a new API service tier that runs GPT-5.6 Sol up to 14 times faster than standard processing. The company says the mode, powered by Cerebras hardware, can generate…
OpenAI has previewed Ultrafast, a new API service tier that runs GPT-5.6 Sol up to 14 times faster than standard processing. The company says the mode, powered by Cerebras hardware, can generate up to 750 output tokens per second. OpenAI is working with select customers to evaluate the tier's performance, according to gadgets360.com.
Every new speed claim from AI labs invites breathless talk of a productivity revolution. Yet raw tokens per second matter little if the output is unreliable or if access is limited to a handful of elite customers. OpenAI's 14x number comes from its own tests and a single hardware partner, Cerebras. The real test is whether this speed holds under real-world load and at what price. Until independent benchmarks and public pricing appear, treat the multiplier as a marketing figure, not a fact.
Source: gadgets360.com
This story was synthesised by AI from the source linked above.