Nvidia reaffirms open source AI strategy following Jensen Huang's exit
Interim CEO Colette Kress confirmed the company will proceed with its open model push and the planned acquisition of Hugging Face.
The short version
- Interim CEO Colette Kress reaffirmed Nvidia's commitment to an open model ecosystem during a recent earnings call.
- Nvidia is on track to acquire Hugging Face for $12.9 billion, its second-largest acquisition to date.
- The Nemotron model family is currently the most widely used U.S. open model on the OpenRouter platform.
- Weekly token volume on OpenRouter saw a 35-fold year-over-year surge, with a 3.5-fold increase in the last three months.
Why it matters
This confirms that Nvidia's shift toward open-source AI is a corporate strategy rather than a personal project of its founder. It signals a continued push for developers to build custom intelligence using Nvidia-supported frameworks.
Continuity in leadership
Nvidia interim CEO Colette Kress recently addressed analysts to clarify the company's direction following the resignation of founder Jensen Huang. During the earnings call, Kress emphasized that the company remains dedicated to fostering a vibrant ecosystem for open models. This move aims to ensure that various industries and countries can develop their own specific types of intelligence.
The transition comes at a sensitive time for the hardware giant, as Huang's departure sparked questions about whether the company would pivot away from open-source initiatives. Kress stated that the goal is to make intelligence more accessible and capable for everyone. By maintaining this stance, Nvidia seeks to provide stability to the developers who rely on their infrastructure.
Despite the leadership change, the core strategy appears to be moving forward without significant modification. Kress noted that open models are currently accelerating the adoption of artificial intelligence across the board. The company intends to continue supporting the tools and models that allow for widespread integration into different business sectors.
The Hugging Face acquisition
A major component of Nvidia's open-source strategy involves the planned acquisition of Hugging Face. The company agreed to purchase the platform for $12.9 billion in September, marking what will be the second-largest acquisition in the history of the firm. Hugging Face serves as a primary hub for open-source AI models and currently supports over 118 million developers.
The acquisition is still on track to proceed despite the recent ouster of Jensen Huang. The platform is used by more than 200,000 companies, making it a critical piece of infrastructure for the AI community. Kress indicated that the goal of the merger is to scale the platform's capabilities to reach a broader audience.
Hugging Face CEO previously expressed sadness over Huang's departure but pledged to continue fighting for open-source AI. Nvidia's plan is to combine its hardware dominance with the platform's massive developer base. This move is intended to make intelligence more open and accessible to the global market.
Growth in open models
Data from OpenRouter, a platform for accessing various AI models, shows a significant trend toward open-source technology. Kress reported a 35-fold year-over-year surge in weekly token volume on the site. A token is a basic unit of text that AI models process to generate responses.
The growth has been particularly sharp recently, with a 3.5-fold increase in token volume over the past three months alone. This suggests that developers are increasingly moving toward open models rather than relying solely on closed systems. This shift provides a clear indicator of how the open-source ecosystem is expanding in real-time.
Nvidia is positioning itself at the center of this growth by providing the necessary tools for these models to run effectively. By supporting high-volume token processing, the company ensures its hardware remains essential for the next wave of AI development. The data suggests that the demand for open intelligence is not slowing down despite the internal leadership changes at Nvidia.
The Nemotron family
One of the specific technologies driving this strategy is the Nemotron family of open-weight models. These models are designed specifically for long-running agentic AI systems and complex reasoning workloads. Open-weight models allow developers to run the AI on their own hardware while still benefiting from the pre-trained intelligence provided by the creator.
Kress highlighted that Nemotron has become the most widely used U.S. open model on the OpenRouter platform. She described it as a foundational tool where developers begin to build their own unique intelligence applications. This success validates the direction Huang championed before his resignation.
The focus on reasoning workloads means these models are built to handle tasks that require multi-step logic. By making these models widely available, Nvidia is empowering a diverse ecosystem of developers. This approach allows companies to build specialized tools without being locked into a single proprietary interface.
Corporate and legal hurdles
While the company moves forward with its AI strategy, it faces significant internal and external challenges. Jensen Huang's resignation followed reports of personal misconduct and allegations regarding his ties to a Singapore-based company. These issues have led to a scheduled congressional testimony for Huang by Thanksgiving.
In addition to the leadership turmoil, the Securities and Exchange Commission is currently investigating Nvidia. The investigation is looking into whether the company's previous disclosures to investors were misleading. These legal and regulatory pressures create a complex backdrop for the company's technological ambitions.
Despite these distractions, the technical roadmap for open-source AI remains the primary focus of the current leadership. Kress's steady communication suggests that the company is trying to insulate its product strategy from the personal controversies surrounding its founder. The focus remains on scaling the platform and making intelligence more accessible to the global market.
What to watch
- The finalization of the Hugging Face acquisition and how it integrates with Nvidia's hardware.
- The outcome of the SEC investigation into Nvidia's investor disclosures.
- How the Nemotron model family performs as more developers adopt it for reasoning tasks.
Try it
Check the OpenRouter platform to see which models are currently trending for developers building agentic systems.
Sources and links
- ReportingTechCrunch: the article we sourced
First reported by TechCrunch. Written by the SI Agent News Desk from the reporting above.Written with AI from the source above. Every number is checked against the source. Spot a mistake? Corrections.
Liked it? Say why
Add a line for other readers.
No account needed. We never show your email. Comments with links, ads or abuse are held and not shown. Your comment shows after a quick check.