Nvidia confirms GPT-6 Astra Ultrafast runs on Blackwell: up to 8x faster token generation

#What happened

On October 1, 2026, NVIDIA published a blog post by Dion Harris titled "How NVIDIA GPUs Help Accelerate OpenAI's GPT-6 Astra Ultrafast." Its key sentence: "GPT-6 Astra Ultrafast, running on NVIDIA Blackwell GPUs, is available now in the OpenAI API and to eligible ChatGPT Work and Codex users."

#The speed claim

Ultrafast offers up to 8x faster token generation than the Astra Standard mode. This 8x figure is NVIDIA's vendor claim about its own hardware, not an independent measurement.

#How it works

NVIDIA attributes the gain to inference optimizations in OpenAI's models that exploit the Blackwell architecture. OpenAI inference lead Philippe Tillet said: "NVIDIA's deep investment in tooling and documentation has enabled us to make our models exceptionally good at programming Blackwell and Rubin GPUs. Astra can turn that knowledge into high-performance kernels that make NVIDIA hardware compelling across the full frontier of latency, throughput and cost. With Astra Ultrafast, that means faster model responses as agents write code, use tools and work through complex tasks." Uday Ruddarraju, OpenAI's chief technology officer of compute, was also quoted in the post.

#Why it matters

Faster token generation shortens coding agents' edit-test-debug cycles and cuts the pause between tool calls, making interactive apps and agentic workflows feel more responsive. The announcement also stakes NVIDIA's claim on the inference layer — OpenAI's own Ultrafast documentation named no GPU, so the Blackwell attribution comes from the hardware vendor.

SOURCES

1. https://aidailypost.com/news/nvidia-ai-infrastructure-powers-openais-gpt — AI Daily Post
2. https://www.orcarouter.ai/blog/gpt-6-astra-ultrafast-on-nvidia-blackwell — OrcaRouter

Verified October 4, 2026.