Palantir’s (PLTR) tie-up with Nebius (NBIS) to provide a full stack of Palantir’s ontology on Nebius’ platform will boost Nebius' revenues. I strongly believe that this is the start of another long-term uptrend in Nebius’ stock price, and this one will carry it past its previous all-time high of $299.86.
Palantir’s (PLTR) tie-up with Nebius (NBIS) to provide a full stack of Palantir’s ontology on Nebius’ platform led to another 7% surge in Nebius’ stock price yesterday. In the last four days Nebius has jumped about 55 points or 25% from $192 to $247. To many Nebius watchers over the past two years, this isn’t surprising, the stock has a 23% short interest so both up and down moves tend to be exaggerated. However, this tie-up gives me a lot of confidence - I strongly believe that this is the start of another long-term uptrend in Nebius’ stock price, and this one will carry it past its previous all-time high of $299.86.
It is my largest holding with about 6.6% of my portfolio, and I’m not going to add more, because it is likely to reach 10% of the portfolio should the momentum continue. But, I strongly recommend a Buy to those who don’t own enough of it
The collaboration delivers a complete “sovereign” infrastructure stack for enterprise customers.
Palantir brings its ontology layer providing software, and application tools, whereas Nebius brings the Compute for training and inference with the full stack of infrastructure from hardware to its AI/token factory. One key aspect is that this whole endeavor will be based on a multitude of open source models instead of being wedded to Anthropic or OpenAI. Nebius lists over 60 such models available to its commercial customers.
The deal will expand its on-premises private cloud business model and will compete directly with the hyperscalers.
Arkady Volozh, Founder and CEO, Nebius had this to say at the Goldman Sachs Technology Conference. I have edited for clarity.
Well, what Palantir calls sovereign is sovereign for the companies. The companies use AI, by using what they produce as a lot of data. This data goes back to the models, improves the models, produces even more data and this loop continues. And the companies have actually today a choice. They can go to commercial models and start feeding their data back to them and improve their models or they can actually use open source, open-weight models on their own -- under their own control, feed data to these models. Keep improving these models for themselves, not sharing that data with the whole world. And actually, the practice shows that if you do this with generally open-weight models are lower quality than commercial models. But commercial models are very universal, very expensive. Open-weight models could be trained with this data in repetitive loops. So that in this specific narrow domain of this enterprise could be trained better to a higher level of intelligence than general models. And there's a lot of cases in our practice, in which it works. Just recently, one of our clients, Shopify,, used an open-weight model, trained it with their own data repetitively, and they achieved the quality which is higher than they had with GPT 5.6. Open model, That's what enterprises should be doing. And by this, they don't share their data. They keep their data. They don't train the whole world, potentially their competitors.
Palantir, NVIDIA and Nebius see this as one of the most important trends in the future of Artificial Intelligence. Simply, enterprise customers making repetitive use of their own data, using open source models to improve the model/outcome/analysis - a self reinforcement virtuous loop that is not shared with commercial frontier models like ChatGPT and Anthropic. I completely agree with this assessment - this data remains private as it should, why should you give away data to train your competitor? Besides, the value addition from an AI solution to a specific vertical or domain is far higher than a general purpose solution. The use of open-weight models in specific narrow verticals/domains would lead to a higher level of intelligence than general models.Palantir and Nebius can make this happen as a better, private and cheaper alternative to Big Data.
From Chief Revenue Officer, Marc Boroditsky, edited for clarity.
Well, as Arkady just shared, there's a very powerful value proposition that we share with Palantir with regards to how we believe enterprise adoption is likely to take place. So working together, combining the full stack capabilities we have and the tooling that they have. And more importantly, the credibility and momentum that Palantir has in the enterprise. We're confident that we're going to be able to jointly develop opportunities that go well beyond the existing market adoption in the enterprise and actually will accelerate our overall enterprise business and contribute to the customer diversification that we have at Nebius.
I think that this is a big big deal and amplified Nebius’ quest to own at least 50% of non-hyperscaler revenue. For the past year, Nebius kept aside capacity for enterprise customers; they always wanted to go to a wider base of clients and this is another big opportunity.
CEO, Arkady Volozh, edited for clarity
Everybody sees what's going on with the prices. It's obvious that demand today is much higher than the real world can build for this demand. And the prices always reflect this imbalance. What's going on is that AI is creating a lot of added value in some way in real life in the industries, be it coding or recently security or office work and many, many other areas which are to come. If we wanted, we could sell all the 2027 demand today, we have much more demand than we can physically serve.
Arkady is obviously in good company when he talks about supply constraints. Jensen Huang and Hock Tan made the same lament in the last two weeks. Arkady claimed demand visibility for 24 months but conservatively is only going to forecast for 18 months - that’s how strong the demand is.
The move to collaborative on-prem private AI token factories/optimization is being played elsewhere. Broadcom (AVGO) along with its software arm VMWare, is also offering similar solutions -tools plus compute services indicating that this is a strong trend as the alternative to the hyperscaler public cloud hegemony of AWS, Azure and Google Cloud. The trend is Nebius’ friend, and I do believe that it will have a first mover advantage.
Another example; Nvidia too will remain in the forefront of AI compute by buying Hugging Face, the largest AI open source platform in the world. It has more than 18 million developers, researchers and creators, and uses more than 3 million models, 500,000 datasets and 1 million applications. More than 200,000 companies use the platform to discover, evaluate, customize and deploy AI.
I believe that open source, open weight models are almost entirely hosted locally on developer systems, and part of the explosion in local LLMs that will gain market share from cloud based models. Nvidia is smart enough to stay ahead of the curb and cover its bets. The Data Center segment is driving revenue growth right now, but the massive expansion can’t go on forever. Nebius has made a very good decision to partner with Palantie, and it adds to their strategy of expanding with the asset light model and bringing on the full stack compute layer without the expensive build outs.
The Palantir deal is part of a broader trend, with Nebius getting significant increases in its pipeline with platform companies that themselves serve enterprise customers. Better still, some are looking at Nebius as their first non-hyperscaler supplier for training post-training and inference across the full AI lifecycle.
Older GPUs are contributing heavily, because many workloads, including RAG infrastructure, text prediction and image generation, can be run more economically on older hardware, giving customers a better TCO, without compromising reliability and performance. They’ve had new customers proactively seek older generation chips
The asset-light model is helping Nebius surmount physical and financial constraints. Here’s how: Nebius’ asset light partners are typically electricity or data center companies that already have land, power and, in many cases, access to cheap financing, but lack the expertise to move higher up the AI infrastructure stack. Nebius provides the data center infrastructure including racks, the software stack and the customer demand to monetize the operation. This is a fairly long and expanding list of customers, with capacity expected to come online from 2027.
Nebius is my largest holding with about 6.6% of my portfolio, and I strongly recommend Buying it. I believe that analysts' revenue forecasts will prove to be conservative.
