08/06/2025
Congratulations to OpenAI for launching two new state-of-the-art, open-source reasoning models optimized for the world’s largest infrastructure. These excellent new models will help developers advance innovation around the world.
The new gpt-oss models were trained on NVIDIA GPUs, optimized across our full stack, and run best on NVIDIA Blackwell and RTX GPUs. The models run on the world’s largest installed base —hundreds of millions of CUDA GPUs — from laptops to data centers and cloud platforms.
“OpenAI showed the world what could be built on NVIDIA AI — and now they’re advancing innovation in open-source software. The gpt-oss models let developers everywhere build on that state-of-the-art open-source foundation, strengthening U.S. technology leadership in AI — all on the world’s largest AI compute infrastructure.” – Jensen Huang, NVIDIA founder & CEO
Read more on our blog: https://nvda.ws/47nGkzg