Nvidia and Stripe Lead Multi-Billion Dollar Sweep of Open-Weight AI Platforms

Major technology companies are aggressively acquiring open-weight AI platforms and model builders to reduce dependence on hyperscalers and hedge against the rise of frontier labs. This strategic pivot sees hardware and fintech leaders integrating the AI software layer to secure their positions within the evolving ecosystem.

Nvidia has reportedly agreed to buy Hugging Face, the largest U.S. developer space for open models, for $12.9 billion. This follows a separate $6 billion agreement with open-weight model builder Poolside, through which Nvidia licensed Poolside's "Model Factory" software and hired 109 of its staff, while also investing $1 billion in the company. Simultaneously, fintech leader Stripe has acquired OpenRouter, a prominent provider of open-weight models for businesses, for more than $7 billion.

The Shift Toward Specialized Intelligence

This acquisition spree comes as frontier labs like Google and OpenAI develop their own inference chips—such as OpenAI's "Jalapeño"—which directly threaten Nvidia's hardware dominance. By owning the platforms where open models are hosted and developed, Nvidia can maintain its influence over how AI is deployed.

At the same time, the industry is seeing a surge in high-volume, repetitive inference tasks, such as customer service, where businesses are seeking cheaper and more configurable alternatives to proprietary models. The scale of this demand is evident in the performance of open-weight infrastructure; Fireworks AI, for example, processes over 40 trillion tokens a day, a volume reported to exceed the API volumes of both Gemini and OpenAI.

The Economic Future of Inference

These moves signal that while proprietary models may currently lead in complex reasoning, the economic future of the industry may rely on "specialized intelligence." This approach suggests a transition toward self-hosted, open-weight models to manage the soaring costs of inference.

Patrick Collison, Stripe cofounder and CEO, noted that tokens have become the central currency for companies building with AI, stating that real-world economic potential will depend on the efficient use of scarce compute resources. This aligns with the vision of Fireworks CEO Lin Qiao, who argues that the future involves every company automatically deploying its own model per specific use case.

What to Watch

As Nvidia and Stripe consolidate their hold on the open-weight layer, the industry will be watching to see if this leads to a more fragmented AI landscape or a new set of consolidated power centers. It remains to be seen how frontier labs will respond to this integration of hardware and software, and whether the shift toward specialized, self-hosted models will successfully erode the market share of closed-source API providers.