OpenAI is previewing a new Ultrafast service tier for GPT-5.6 Sol that can run the model up to 14 times faster than standard processing, giving developers a much faster option for workloads where response time matters.
The new mode launches first through the OpenAI API and uses Cerebras infrastructure to generate up to 750 output tokens per second, which gives businesses more room to use advanced AI in live applications without waiting as long for results.
Don’t miss the best of The Mac Observer
Set us as a preferred source and our Apple reporting ranks higher in your Google Search results and Discover feed — one tap, no account changes.
OpenAI introduced GPT-5.6 Sol in June alongside the balanced Terra and speed-focused Luna models, while the wider GPT-5.6 family became broadly available in July across ChatGPT, Codex, and the API.
The company expects Ultrafast mode to support workloads such as voice applications, customer support, commerce, developer agents, financial research, and security response, where teams often need useful answers within seconds.
OpenAI also says its developers have used the faster tier to analyze logs and traces during incidents, while some research tasks that previously ran overnight can now support several iterations during a normal workday.
Access remains limited to selected customers while OpenAI studies how the higher speeds affect real-world products and expands capacity. Businesses can join the Ultrafast waitlist by providing details about their workload, latency needs, and expected usage.
Discussion