Roman Chernin

My [optimistic] thinking: Anthropic’s insane revenue growth and the huge number of tokens it generates show what will happen next with open-source / alternative model tokens. Start with frontier models → grow usage → collect data → tune cheaper models -> get margins.

People I need to meet next time in the bay area?

That’s how we do it:)

Today, we're announcing that Eigen AI is joining Nebius (NASDAQ: NBIS). From day one, our mission has been Artificial Efficient Intelligence — building the world's most efficient engines for generating intelligence. Together with Nebius, we're working toward the best AI cloud, uniting Eigen's full-stack model and inference software, ranked #1 on Artificial Analysis for inference speed, with Nebius's global hardware and infrastructure footprint, so any developer or enterprise can run the best models at the best price, with no capacity ceiling. After close, Eigen's optimization stack will be integrated directly into Nebius Token Factory. The entire Eigen AI team is joining Nebius in full, establishing Nebius's engineering and research presence in the San Francisco Bay Area. To our customers, our team, our investors at Tectonic Ventures, E14 Fund, Uncorrelated Ventures, and AGI House Ventures, our angel investors, advisors, mentors, and supporters — and to the Nebius team for the conviction and partnership — thank you. The mission doesn't change. The leverage behind it does. Ryan Hanrui Wang, co-founder and CEO of Eigen AI, said: “We’re proud to join Nebius and work alongside the Token Factory team to push the boundaries of inference performance. Nebius has built a world-class AI cloud with a deep engineering culture that perfectly aligns with our own. Together, we are removing the friction of AI model customization and deployment so developers can run models reliably in production without managing the underlying infrastructure.” Full announcement at: eigenai.com/blog/eigen-ai-…

265 tokens/sec on Kimi K2.6. @Eigen_AI_Labs leading the pack ⚡ Eigen 🤝 Nebius Token Factory

Everybody’s talking about data gravity. But the reality is that inference generates tons of new data. At @nebiustf, we’re taking a step toward helping builders collect, curate, and use that data as they build applications at scale. nebius.com/services/token…

Data Lab is the workspace where production logs and existing data become reusable datasets for model improvement. Explore inference data, filter real-world signal, connect files from S3-storage, and move curated datasets into post-training without stitching tools together.

$NBIS cofounder @romanchernin explains that by 2026, every product will be enabled by tokens and why infrastructure for alternative models is critical: "In 2026 every product that you build is enabled with these tokens, so-called tokens—so, intelligence. You can get tokens from frontier models... or you can get tokens from alternative models, whether it be open source or alternative model builders." "We help people to work with these alternative, like open source and specialised models... because when you go to Anthropic you get an API; it just works for you. But if you go and you want to work with GPUs... you can download weights from Hugging Face... but the reality is it will not work, or at least it will not work at scale." "It will not work at scale with the economics and reliability that you expect. It's a quite complex engineering task to run a large workload and extract every cent, every percent of economics from this infrastructure, and that's why products like Token Factory have a right to exist".

@RayT168 Andrey can)

@romanchernin Is it true Dan can build data centers with his eyes closed?

How we build from Dan, Nebius CTO nebius.com/blog/posts/the… We are hiring.

@Mhr1036 @tbpn Break them sometimes

@tbpn @romanchernin Move fast, dont break things

$NBIS cofounder @romanchernin describes how their recent acquisitions of Eigen AI and Clarifai were all about speed, incredible talent, and acceleration: "The philosophy is very simple. We need to build so many things, and we need to move so fast, that we're always looking for people who can accelerate us. It should be exceptional talent, and/or something that has a great adoption." "Our two recent acquisitions [were] two teams that work on inference optimization. A big part of our business is how efficiently we convert GPUs into tokens. And these two teams — Eigen AI and Clarifai — one is focused on model optimization, the engine of inference. How you run specific models and all the techniques around spec decoding, quantization, and so on." "And the other is system optimization. All the routing, KV caching, and orchestration across the big cluster of compute and so on." "We have a very strong internal team working on inference. But we felt that we needed to move faster, bring more capabilities. Because the market is so fast."

Honored that such a talented team has decided to continue building as part of Nebius! Let's go!

We are live with Q1 results

Inference efficiency is a race, not a static snapshot. To maintain leadership over time, you need the best racers on your team who can keep pushing the frontier and consistently deliver the best results for customers.

nebius cooks beyond all other cooks

AI infrastructure demand isn't slowing down — and @nebiusai is building to match it. Arkady Volozh, CEO of Nebius, gives his perspecitve on the company's full-stack platform, the NVIDIA partnership, and what it takes to serve customers at scale. Watch now ⤵️

Agentic inference isn’t a token speed race. It comes from optimizing the whole loop: model, memory, orchestration and compute That's where @Eigen_AI_Labs + @clarifai + @nebiustf come in!

We are dead serious about building the best inference platform. I can’t wait to see what the teams of @nebiustf, @Eigen_AI_Labs and @clarifai deliver together.

Welcome, Clarifai team – Nebius will take care of you nebius.com/newsroom/nebiu…