Amazon Web Services (AWS) and OpenAI have signed a multi-year partnership to strengthen large-scale artificial intelligence (AI) operations using AWS’s cloud infrastructure. The $38 billion agreement, which will run for seven years, gives OpenAI access to hundreds of thousands of NVIDIA GPUs, with room to scale up to tens of millions of CPUs as its workloads grow. AWS will underpin OpenAI’s expanding agentic systems, drawing on its track record in running secure, high-performance AI clusters that already exceed 500,000 chips. The AWS and OpenAI partnership combines AWS’s cloud expertise with OpenAI’s leadership in generative AI to keep powering ChatGPT for millions of users worldwide.
Right now, demand for high-performance computing is climbing fast. It’s being fueled by the rapid growth and complexity of AI technologies. Under the new agreement, OpenAI will start using AWS compute resources right away, with full capacity expected to come online by the end of 2026. The companies expect to expand further through 2027 and beyond to meet OpenAI’s growing computational demands.
At the core of the partnership, AWS is building an infrastructure made to handle demanding and complex AI workloads. It uses Amazon EC2 UltraServers to connect NVIDIA GB200 and GB300 GPUs within a unified network, ensuring low latency and smooth coordination across systems. This setup will handle a broad mix of OpenAI workloads, from training next-generation models to running inference for ChatGPT. It’s also built to scale smoothly as those demands continue to grow.
“Scaling frontier AI requires massive, reliable compute,” said Sam Altman, co-founder and CEO of OpenAI. “Our partnership with AWS strengthens the broad compute ecosystem that will power this next era and bring advanced AI to everyone.” Matt Garman, CEO of AWS, added, “As OpenAI continues to push the boundaries of what’s possible, AWS’s best-in-class infrastructure will serve as a backbone for their AI ambitions,”
The collaboration between AWS and OpenAI builds upon earlier joint efforts to bring OpenAI’s technology to AWS customers. OpenAI’s open weight foundation models, now available through Amazon Bedrock, have quickly gained traction among enterprise users such as Bystreet, Comscore, Peloton, Thomson Reuters, Triomics, and Verana Health, who apply these models to scientific analysis, coding, and agentic workflows.