OpenAI announced on Thursday (13) the Ultrafast mode, a new feature of GPT-5.6 Sol that operates at 14 times the standard processing speed.
In a blog post, the company said the feature generates up to 750 output tokens per second. Tokens are text fragments that the model uses to interact with humans.
OpenAI said in the post that, until now, getting real-time speed required choosing a smaller or more specialized model. "Ultrafast points to progress in a new direction: more useful work per second."
The company cited incident response, customer service, financial market analysis, and e-commerce as application areas.
Availability
Ultrafast is powered by a partnership with chipmaker Cerebras. In preview, the feature is available to a restricted group of customers, and OpenAI said it will expand access as capacity grows.
Rival Anthropic also launched a fast mode for Claude, but without delivering the same level of speed as Ultrafast.
