NVIDIA gives away an AI giant for free: the secret to billion-dollar profits on an open model
On June 4, 2026, NVIDIA released the Nemotron 3 Ultra — the largest open AI model in the Nemotron 3 lineup. Unlike closed flagships such as ChatGPT or Claude, this model is fully available under a free license: weights, training data, and methodologies can be downloaded, fine-tuned on your own data, and run on your own infrastructure. The bet is not on record-breaking intelligence, but on openness, efficiency, and full control.
Architecture: Three in One
The Nemotron 3 Ultra is not just a "scaled-up transformer." It is based on a hybrid architecture combining three approaches: Mamba-2 layers, Attention, and Latent MoE (Mixture of Experts). Mamba-2 processes long texts quickly and efficiently — costs grow linearly, not exponentially. Attention accurately handles large context volumes. And Latent MoE compresses data before passing it to experts, allowing each to work narrowly and precisely without unnecessary computation.
The model has approximately 550 billion parameters in total, but only about 55 billion are activated for processing each token. This gives it 5–6 times higher throughput than comparable models and roughly 30% lower task costs. The context window is 1 million tokens, with a speed exceeding 300 tokens per second.
Strategy: Free AI as a Driver for Hardware Sales
The main value of the release is not the model itself, but the ecosystem NVIDIA is building around its hardware. The logic is simple: anyone running Nemotron is almost certainly doing so on NVIDIA GPUs, fine-tuning it with NVIDIA tools, and deploying it on NVIDIA software. Openness here is not charity, but a way to steer developers back to purchasing the company's hardware.
With a market capitalization exceeding $5 trillion, training Nemotron 3 Ultra — which likely cost hundreds of millions of dollars — is a nearly negligible expense for NVIDIA. GPU sales more than cover the research costs, allowing the company to give the model away for free and still earn more than closed competitors do with paid access.
The political context adds further weight to the release. An open American model can be inspected, modified, and run on one's own servers — making it attractive for countries building independent national AI, from Europe to Southeast Asia. Such a model cannot be remotely disabled, which is especially valuable amid recent restrictions surrounding closed models.
Where the Model Falls Short and What's Next
Despite its strengths, the Nemotron 3 Ultra is not the smartest model on the market. In the independent Artificial Analysis Intelligence Index, it scored 48 points — the best result among open US models, but trailing leaders like Kimi K2.6 (54 points) and DeepSeek. Open models, according to analysts, lag behind closed ones by three to seven months.
But this gap, in my view, matters less and less if an open model is simply sufficient for real-world tasks. A bank deploying Nemotron 3 Ultra to process loans on its own servers doesn't need flagship-level intelligence — it needs a model that can be fine-tuned on proprietary data, kept within a secure perimeter, and not hand over confidential information to third parties.
My conclusion: NVIDIA's bet on efficiency rather than benchmark records may prove more farsighted. In mass AI adoption, the cost of running a model comes to the forefront, and one that is almost as capable but five times cheaper wins in real-world deployment. I expect the open ecosystem to only strengthen: NVIDIA has the resources, motivation, and distribution channels to release increasingly powerful open models faster than any other company.