GPT-6 Astra Ultrafast: OpenAI's 8x faster response mode now live
OpenAI has released GPT-6 Astra Ultrafast, running on NVIDIA Blackwell GPUs, delivering up to 8x faster token generation than Astra Standard, now available in the OpenAI API and to eligible ChatGPT Work and Codex users.
OpenAI has introduced GPT-6 Astra Ultrafast, a mode running on NVIDIA Blackwell GPU infrastructure that delivers noticeably faster responses than the Astra Standard mode. According to an October 1, 2026 announcement on the NVIDIA Blog, the mode is available now in the OpenAI API and to eligible ChatGPT Work and Codex users.
Who can use it
- Developers via the OpenAI API.
- Eligible ChatGPT Work users.
- Eligible Codex users.
Why it's faster
Astra Ultrafast is accelerated through continuous inference optimizations OpenAI applies on top of the NVIDIA Blackwell architecture, offering up to 8x faster token generation than Astra Standard. OpenAI says this speed shortens the write-code, use-tool, check-result loop for coding agents and makes interactive applications feel more responsive.
How to access it
Developers can start using GPT-6 Astra Ultrafast today through the OpenAI API. For access details, pricing, and implementation, OpenAI points users to its separate Ultrafast guide on the developer site; the announcement itself does not list specific pricing.