Access Nebius through Foundry Cloud Platform

Partnering closely with AI infrastructure companies in a manner focused on interoperability, reliability, security, and scalability is critical to providing our customers with the flexibility and control they desire, both in terms of user experience and economics. With that, we’re excited to announce a strategic partnership with Nebius, a global AI infrastructure company, to bring the simple, flexible compute provisioning of Foundry Cloud Platform to Nebius’ infrastructure, powered by NVIDIA H100, H200, and Blackwell GPUs. 

Our integration with Nebius enables Foundry customers to access existing Foundry capacity as well as new Nebius AI Cloud regions through the Foundry Console for experimentation, large-scale training, and production inference workloads. We’ve rigorously tested and benchmarked Nebius’ infrastructure and our integration to ensure that all compute provisioned through Foundry Cloud Platform continues to deliver the performance and simplicity that Foundry customers rely on. 

Accessing Nebius GPU capacity through Foundry

Starting today, NVIDIA H100 and H200 GPU instances provided by Nebius are available via custom reservations across three new regions in Foundry Cloud Platform: eu-north1, eu-central1, and eu-west1.

For even heavier workloads, we are also excited to bring NVIDIA GB200 NVL72 and NVIDIA HGX B200 clusters to Foundry for the first time through our partnership with Nebius. Blackwell architecture delivers major performance and power efficiency gains and includes significantly increased on-device memory, leaving more headroom for large models like Llama 4 Maverick or DeepSeek-R1.

To make custom NVIDIA H100 and H200 GPU reservations or request upcoming NVIDIA Blackwell GPU capacity, contact your Foundry account team or email sales@mlfoundry.com. 

What’s next

In addition to new compute types, we are working on additional technical integrations with Nebius to make it even easier for Foundry customers to provision exactly the compute they need. 

“We’ve architected Foundry Cloud Platform and development of the surrounding ecosystem with an emphasis on being able to support developers throughout their development and production journey. In Nebius, we’re excited to partner with a leading AI infrastructure company that similarly has a customer-first mindset, and we look forward to bringing the best of Nebius and Foundry to customers.”

- Raghav Kohli, Head of Strategic Partnerships, Foundry

“We at Nebius AI Cloud consider flexibility a key quality when it comes to providing our infrastructure for model training, fine-tuning, and inference. This includes flexibility in how you can access compute across different cloud regions. We are thrilled that long-term Foundry customers and experts familiar with the platform can now provision resources directly in their native environment — with the agility, performance, and support they need.”

- Laurelle Roseman, Vice President of Channel & Alliances, Nebius