NVIDIA Says Blackwell GPUs Help GPT-6 Astra Ultrafast Reach Up to 8× Faster Token Generation
NVIDIA says GPT-6 Astra Ultrafast runs on NVIDIA Blackwell GPUs and is available through the OpenAI API and to eligible ChatGPT Work and Codex users, offering up to 8× faster token generation than Astra Standard.

Full Report
NVIDIA says GPT-6 Astra Ultrafast runs on Blackwell GPUs and is accelerated through ongoing inference optimizations using OpenAI's own models. The mode is available through the OpenAI API and to eligible ChatGPT Work and Codex users.
According to NVIDIA, Ultrafast can deliver up to 8× faster token generation than Astra Standard, with benefits for code generation, tool use and interactive applications. NVIDIA also says OpenAI is using its own models to help optimize inference software running on NVIDIA GPUs, with the goal of improving response speed and infrastructure productivity over time.
Why This Matters
Faster token generation can reduce latency across coding agents, tool calls and interactive workflows. The “up to 8×” figure is NVIDIA's stated performance claim and should be interpreted in the context of its measurements and OpenAI's actual access conditions.
Evidence
Start with the primary evidence, then review supporting sources and context.
Primary Evidence
1How NVIDIA GPUs Help Accelerate OpenAI’s GPT-6 Astra Ultrafast
Dion Harris
“GPT-6 Astra Ultrafast, running on NVIDIA Blackwell GPUs, is available now in the OpenAI API and to eligible ChatGPT Work and Codex users. Accelerated by inference optimizations through OpenAI’s models that tap into the capabilities of the NVIDIA Blackwell architecture, Ultrafast offers up to 8x faster token generation than the Astra Standard mode.…”
Recommended signals
Recommended by brand, topic, and recent related coverage