Nebius Inflection: AI Expert Forum

The first phase of AI was about models and infrastructure. Making them work reliably, efficiently, and at scale. That’s where we started at Nebius. But the next phase is different. Models are improving, developers are becoming AI natives, and agents are challenging the infrastructure status quo. If the first wave of AI was defined largely by access to GPUs and scale, the second one will be about enabling builders to iterate fast with production consistency.

Agentic systems will run continuously, evolve from real-world usage, call tools, route across models, and operate at a level of complexity that today’s infrastructure was not originally designed to support.

Raw compute or a long list of cloud services will not be enough. Developers will need fast access to predictable infrastructure, low-latency inference and improved unit economics, model and agentic flexibility for a simple path from prototype to production.

We are already building our global AI cloud for this new era. And on June 9th in San Francisco, we’re bringing together a small, invitation-only group of the people that will define what that next phase looks like.

Builders, operators, investors and thought leaders, joining an honest conversation about what’s working, what’s breaking, and what needs to exist. You will hear from the customers and partners building with us, get into the details of what the agentic phase actually demands and be among the first to hear what we’re announcing, with direct access to the Nebius team.

So we can figure out what comes next, together.