Blog

Discover how Nebius AI Studio enables you to integrate the famous Llama 3.1 405B large language model in your applications.

We have developed a new version of Nebius platform that we believe will serve your needs even better. It is already being tested by our in-house LLM R&D team and a number of clients. Now, we are rolling it out to everyone.

The current expansion phase will enable us to place up to 60,000 GPUs at the Mäntsälä location alone.

In September, we announced the opening of our French region, launched the world’s first open-source K8s operator for Slurm as well as Nebius AI Studio, a product for GenAI builders — and there’s more.

During the upcoming maintenance, we expect a downtime for our VPC (Virtual Private Cloud) service, resulting in possible short disruptions to network connectivity.

Our new platform not only allows application builders to use generative AI without significant effort but does so at a fraction of the cost. Our goal is to democratize access to cutting-edge AI technologies, enabling businesses of all sizes to innovate and compete in the AI space.

We’ve built an open-source Kubernetes operator that runs and manages Slurm clusters as Kubernetes resources. From this article, you’ll find out how the community tackled this task previously and the details of the architecture of our solution, which we named Soperator.

We’ve developed Soperator, a Slurm-based workload manager with a modern and simplified user experience for ML and HPC cluster orchestration.

Machine learning inference tests a model against real-world unseen data. Understand what is inference, and how models are deployed in production.

One of the experts from our LLM R&D team, Alexander Golubev, recently took part in the Kaggle competition hosted by the Large Model Systems Organization (LMSYS). Here are his impressions.

Explore our comprehensive guide to automated machine learning (AutoML) and find out why it’s changing data science and making machine learning available to everyone.