OpenAI's new Ultrafast mode makes GPT-5.6 Sol run at 14 times normal speed
A new preview mode promises dramatically faster responses from OpenAI's most powerful model. Here's what that actually means and who can get it.
Key points
- OpenAI launched a new mode called Ultrafast for its GPT-5.6 Sol model on 13 August 2026.
- Ultrafast generates up to 750 tokens (pieces of text) per second, which OpenAI says is 14 times faster than standard processing.
- The feature runs on chips from Cerebras, a specialist AI chip company, under an existing partnership with OpenAI.
- Ultrafast is currently in preview and available to a limited group of business customers only.
- Anthropic's Claude chatbot has a similar fast mode, but OpenAI claims its speed figures are higher.
If you've ever typed a question into ChatGPT and drummed your fingers while it thought, OpenAI has noticed.
On Thursday the company announced Ultrafast, a new speed mode for GPT-5.6 Sol. The promise: responses generated up to 750 tokens per second, 14 times faster than usual. A token is a small chunk of text, roughly three quarters of a word, that a large language model (the technology behind chatbots like ChatGPT) produces as it builds a reply. Our 6 August story on GPT-5.6 Sol for paying subscribers is worth a read for context on how the model itself has been evolving.
Who is this actually for?
Not the average ChatGPT subscriber, at least not yet. OpenAI's rolling out Ultrafast as a preview for a small group of business customers, with plans to expand as capacity grows.
The pitch is squarely at corporate workflows where speed genuinely matters: incident response (catching a server outage before it spreads) and customer service queues, financial market analysis, e-commerce.
"Until now, getting real-time speed typically meant choosing a smaller or more specialized model," OpenAI wrote in its announcement. "Ultrafast points to progress in a new direction: more useful work per second."
That quote matters. Faster AI has historically meant simpler AI. Ultrafast, if the claim holds, would let businesses get speed without trading away quality.
What makes it so fast?
Cerebras, a company that builds specialised processors designed for heavy AI workloads rather than the general-purpose chips in most computers, is powering the mode under an existing partnership with OpenAI.
Anthropic, the company behind the Claude chatbot, has its own fast mode. It doesn't match OpenAI's 750-tokens-per-second claim, though direct comparisons depend on the task.
What does this mean for everyday users?
Nothing to do right now. Ultrafast is a business-tier, invite-only preview. Regular ChatGPT users won't see it in their settings today.
If you work somewhere that already uses OpenAI's API (the programming interface companies use to plug ChatGPT into their own products), watch for the wider rollout. The broader story here is that response times across AI tools are improving fast, and competition between OpenAI and Anthropic will keep pushing that.
Common questions
Will Ultrafast cost more than standard ChatGPT?
OpenAI hasn't published pricing for Ultrafast yet. Speed-optimised tiers in AI services typically carry a higher per-use cost for businesses, so expect a premium once it leaves preview. Our earlier piece on how unpredictable AI token costs are catching companies off guard explains why that bill could be harder to forecast than it looks.
Does Ultrafast change what GPT-5.6 Sol can do, or just how fast it does it?
Speed only, based on what OpenAI has shared. The underlying model is the same; Ultrafast is about delivering answers faster, not adding new abilities.
Is my data handled differently in Ultrafast mode?
OpenAI hasn't announced separate privacy terms for Ultrafast. The same data-handling rules that apply to GPT-5.6 Sol in standard mode should apply here, but businesses should check their enterprise agreements before rolling it out.
