OpenAI releases Ultrafast mode for GPT-5.6 Sol, makes the AI 14 times faster

OpenAI is making GPT-5.6 Sol faster with a new Ultrafast mode. The company says that the AI model can be as 14 times faster in this mode, allowing you to get more work done.

In Short

• OpenAI previews Ultrafast mode for GPT-5.6 Sol
• Company says it makes Sol 14 times faster
• It is available to limited businesses for now

AI models are getting better almost every day. OpenAI’s GPT-5.6 Sol is one of the most advanced AI models you can use today. But the problem with frontier models is that they make time to process requests. And now OpenAI says it has a solution with a new Ultrafast mode for Sol, making the model up to 14 times faster.

In a blog post, OpenAI announced Ultrafast as a new service tier for Sol. OpenAI says that Ultrafast runs “GPT5.6 Sol up to 14 faster than Standard processing.” The company says that Ultrafast will allow you to do “more useful work per second.”

This move comes at a time when US AI labs like OpenAI are facing stiff competition from Chinese open-weight models – that can be run locally on-device – such as Moonshot’s Kimi K3.

While OpenAI’s competitors, including Anthropic, have also launched faster options such as Claude’s fast mode, OpenAI said Ultrafast is designed to deliver frontier-level performance at much higher speed.

Ultrafast is for power users

According to OpenAI, Ultrafast can generate up to 750 output tokens per second. For those unaware, tokens act as a basic unit of measurement for AI models – the more work you do, the more tokens you consume.

Ultrafast is designed for users who have wanted ChatGPT to be quicker without having to switch to smaller AI models. OpenAI says that the mode is powered by AI hardware company Cerebras that works on ultra-low latency AI inference.

According to the company, Ultrafast can be used for tasks where speed is essential. It can also help analyse application logs, recent code changes and engineer reports while an outage is still unfolding. Or assess transactions and suspicious activity while conditions are changing and resolve customer issues in real time.

OpenAI’s own developers have been using Ultrafast during incidents to read logs, analyse traces, synthesise conversations, identify next checks and help prepare or validate a fix. Ultrafast has also helped tighten research loops at OpenAI that earlier ran overnight.

Who can use Ultrafast?

The company is releasing the feature in preview through API for now. During the preview period, OpenAI is working with an initial group of customers to study where the higher speed changes real-world products the most.

Ultrafast is being tested with companies across coding, commerce, financial research, support and other interactive applications, including Jane Street, Podium, Basis and Rogo.

Businesses can join the Ultrafast waitlist by sharing details such as workload, latency requirements and expected usage, as OpenAI evaluates the early rollout and prepares to widen access.

OpenAI has not disclosed the token costs for Ultrafast as of now.

- Ends