Nebius: Killed By Chinese Models?

$NBIS Q4 2025 Earnings Update

The content of this analysis is for entertainment and informational purposes only and should not be considered financial or investment advice. Please conduct your own thorough research and due diligence before making any investment decisions and consult with a professional if needed.

The most important sentence in the whole Nebius earnings call was this:

Demand from enterprises and AI native customers continues to outpace supply, allowing us to sell future capacity well in advance.

This is what you want to hear from any AI infrastructure company. We heard the same kind of language from Microsoft, Google, and Amazon. But when a neocloud says it, it matters even more, because it means the market is not “nice to have”, it is “I need it yesterday”.

Table of Contents

• Q4 2025 Update

• Lorenzo2cents (L2C) take aways and performance

• Business Ontology Framework by L2C

  • Business Ontology

  • 4D Valuation Model

• L2C portfolio strategy

Q4 2025 Update

Nebius reiterated a very bold target: an annualized revenue run rate of $7B to $9B by the end of 2026. And management sounded more confident than before.

Over the last few months, our conviction in this range has become stronger. Why? Because we exceeded the high end of our 2025 ARR guidance and showed more than $1.2 billion ARR, because we already contracted more than 2 gigawatts of capacity and are on track to exceed 3 gigawatts this year, because we have already delivered all of our capacity for the Meta contract, because we are on track to deliver the capacity for Microsoft through the course of 2026, exactly as planned. And lastly, the demand for our AI cloud continues to be strong.

So yes, demand is strong.

But the real story, in my opinion, is why demand stays strong and why Nebius might become a “go-to hyperscaler” for the next generation of AI-native companies.

And here comes my main narrative: Chinese open-source models are a big, underrated tailwind for Nebius.

The “token factory” thesis (and why open source makes it stronger)

I think too many people still look at AI compute as a generic commodity. “GPUs are GPUs.” That is the shallow view.

The deeper view is: the winners will be the companies that can turn GPUs into the cheapest, most reliable, most scalable token factory.

Tokens are the new unit of production:

• Not seats.

• Not licenses.

• Not “cloud instances” as we think about them today.

If the next decade is built on agents, copilots, and AI-native apps, then the world will buy tokens, not software seats.

Now, what changes the economics of tokens?

Models.

And specifically: high-quality open source models that keep getting better.

This is where China matters. Models like Kimi K2.5 and Minimax M2.5 are part of a trend: open source is not “second best” anymore. It is getting dangerously close to the frontier, sometimes it is better for specific tasks, and it is improving fast.

That creates a massive tailwind for companies like Nebius because:

1. Open source reduces pricing power at the model layer

   If anyone can host strong models, then the “model owner” does not capture all the value.

2. Open source shifts value to infrastructure + delivery

   If the model is not locked behind a single API, customers care more about:

   • performance per dollar

   • latency

   • uptime

   • scaling

   • security and enterprise controls

3. Open source increases experimentation

   When costs drop and restrictions drop, more startups build, more teams deploy, and usage grows faster.

And Nebius is positioned exactly where this value migrates: the production layer.

Nebius can say: “Bring your model. We will run it better and cheaper.”

That is token factory thinking.

“Robust performance at the cheapest price” is not a slogan. It is a strategy.

When models become more open, the competitive game becomes brutal:

• the buyer has more options,

• switching costs go down,

• and margins compress for undifferentiated providers.

So the infrastructure provider needs a real edge:

• better software stack,

• better orchestration,

• better reliability,

• and cost discipline.

Nebius is not trying to be “a GPU landlord”.

They are trying to become an AI hyperscaler.

That is a very important difference.

AI-native customers are becoming real enterprises (and that is the pipeline)

The second most important thing I heard in this call was this part:

AI start-ups are quickly evolving to real enterprise scale customers with revenues from their products resonating with real customers. Their demand quickly grows from hundreds of GPUs to tens of thousands. This is already a range the market saw from the biggest customers last year.

And then they give examples:

There’s a lot of customers from the sector, starting with Cuckoo, Cursor, Rodo, Higgsfield, Photoroom, Genesis Molecula, different sectors, and they all have their first traction and they’re becoming -- as we speak, becoming real, real companies. Actually, it’s the future enterprises.

This is the biggest long-term reason I am invested in Nebius.

The next mega-companies will be AI-native. Many of them are being born right now. Their demand curve is not linear. It is explosive.

And if Nebius becomes the “default cloud” for these winners, the compounding is insane.

---

---

Nebius is not “compute for hyperscalers”. Nebius wants to be the hyperscaler.

I repeat this because it is the core of my thesis.

Most neocloud companies survive by selling big deals to hyperscalers. They are basically a supplement, a pressure valve, or a short-term capacity partner.

Nebius said something that shows a different mindset:

Based on the traction we are experiencing and our extensive research on our total market opportunity, we are leaning into the verticals we’ve already laid out, health care, life sciences, media and entertainment, physical AI and retail, which give us ample runway to capture share and grow our business. While we are happy to service large strategic customers like hyperscalers, we will remain opportunistic with such large deals as we look to balance the opportunity with the long-term positioning we plan to achieve with our AI cloud

This is exactly what you want.

Hyperscaler deals are not the “oxygen” for Nebius. They are “gold opportunities” if the economics are good and if they help scale faster.

That’s a huge difference versus competitors.

Building the software moat: Tavily acquisition and the “developer platform” direction

If Nebius wants to be a hyperscaler, it cannot only be hardware and data centers.

It needs software. It needs a platform developers love.

This is why the Tavily acquisition is interesting.

Tavily is agentic search company. They connect AI agents to the web. And it very much fits in our strategy to become and to be the platform where all the AI developers from start-ups and enterprises building their AI applications and agents.

Then:

In addition to our organic roadmap, we take a product-led approach to expanding our tech stack and tech talent with M&A. The acquisition of Tavily, announced this week, expands our AI platform with market-leading agentic search capabilities. This gives Nebius customers the critical infrastructure they need to build autonomous agents that can navigate the web, verify facts, and execute complex real-world tasks, while eliminating the need for developers to patch together disparate vendors.

I like this because it is consistent with the token factory idea.

Tokens are not useful alone. Tokens are useful when they become:

• agents that act,

• apps that sell,

• workflows that replace labor.

Tavily is part of the “enablement layer” that makes Nebius sticky.

Enterprise adoption: Aether 3.1 and the roadmap

Nebius is also pushing hard on enterprise readiness:

We aim to make it easier for enterprises to adopt AI at scale, with features that enable easier integration into enterprise environments, deliver security and compliance, and enable cross-functional operations.

They released Aether 3.1 in December 2025:

• more transparency,

• greater controls,

• better enterprise-grade security.

And they claim a robust pipeline for 2026:

• integration with on-prem and multi-cloud,

• stronger security and compliance,

• operations across teams,

• and vertical-specific solutions (healthcare, physical AI, media, retail/e-commerce, financial services).

This matters because open source is not only a startup story.

If open source models like Kimi and Minimax keep improving, enterprises will want:

• flexibility,

• control,

• and deployment options.

Nebius can meet them there.

Lorenzo2cents (L2C) take aways and performance

My core takeaway from this quarter is simple:

1. Demand is outpacing supply.

2. Nebius is scaling capacity fast.

3. AI-native customers are becoming enterprises.

4. Nebius is building a software moat with acquisitions like Tavily.

5. And Chinese open-source models are an underrated tailwind because they accelerate tokenization and commoditize the model layer, pushing value into delivery.

If Nebius becomes the best “token factory” for robust performance at the cheapest price, then it is not “a neocloud”.

It is an emerging hyperscaler.

And that is exactly why I’m here.

As always, here’s the Deep Dive To Date (DDTD): Stock performance since my first deep dive and when I bought in on July 28, 2025, at $51.63.

+90% DDTD

From here on, the content is restricted to L2C Premium Members, folks who’ve chosen to unlock this toolkit and support my independent research:

• Business Ontology Framework by L2C

  • Business Ontology: My core blueprint for modeling and tracking company performance at every level.

  • 4D Valuation Model: the valuation tool I use to value all my investments (Fair price is useless)

• L2C Portfolio Strategy: My portfolio allocation and strategy in details

• L2C Portfolio access & trades alerts: Real-time views into my holdings, plus instant notifications on buys, sells, and shifts.

Business Ontology Framework by L2C

The Business Ontology is a framework I built after tearing apart several tech companies from the ground up—breaking them down to their basic parts and piecing together a real thesis on what drives them. Think of it as a map of a company’s soul. It’s a tight set of core indicators—tailored to each business—that show if it’s heading the right way, no matter what the stock price says. These aren’t your basic stats like P/E ratios or revenue bumps you grab from Yahoo Finance. They’re deeper, sharper, and linked straight to the thesis I’ve cooked up on how the company makes value and fights in its market.

The framework boils down to two big pieces:

1. The Business Ontology—This checks if the company’s worth buying into or hanging onto.

2. The 4D Valuation Model—This gives you a 4D roadmap to guide your own calls on how much to put in (allocation).

Now, let’s dive into Nebius and this quarter’s update.