- Nvidia is reportedly in talks to acquire Hugging Face, reflecting a broader trend of open-weight AI companies becoming prime acquisition targets in the Valley.
- Stripe acquired OpenRouter for more than $7 billion, further emphasizing the trend of significant investments in open-weight AI startups.
- Nvidia's acquisition of Hugging Face would allow it to gain access to a mass of users it can drive to its chips and standards.
- Recent acquisitions highlight the trend: Nvidia struck a $6 billion agreement with Poolside, an open-weight model builder, and Stripe acquired OpenRouter for more than $7 billion.
- Adoption of open-weight models is still small but growing: 6% of companies use them, according to Ramp, and 2% of software engineers, according to a tool for developers.
Nvidia's potential acquisition of Hugging Face underscores a significant shift in the AI landscape, where open-weight models are becoming increasingly valuable.145
Following a recent $6 billion deal with Poolside, Nvidia aims to reduce its reliance on major hyperscalers and gain access to a broader user base.
Currently, only 6% of companies utilize open-weight models, according to Ramp, indicating substantial growth potential.6
Nik Albarran, AI product lead at Jellyfish, noted that these models are particularly beneficial for businesses with repeated inference workloads, such as customer service applications.
Patrick Collison, co-founder and CEO of Stripe, emphasized the economic potential of AI, stating, “Tokens are the central currency for companies building with AI.”2
As prices from frontier labs rise, more companies may consider investing in self-hosting models.
Lin Qiao, CEO of Fireworks, highlighted the importance of in-house research, suggesting that every app company should develop specialized models tailored to their use cases.
The trend of acquiring open-weight startups, as seen with Stripe's recent acquisition of OpenRouter for over $7 billion, reflects the growing capital influx into this sector, which is based on providing open access to AI technologies.
“The deal would give Nvidia control of the largest U.S. developer space for open models, driving users to its chips. Adoption remains small but growing, with just 6% of companies using open-weight models, per Ramp, and Fireworks processing 40 trillion tokens daily.”










