# Nvidia CEO Jensen Huang Backs Both Open and Closed AI Models Nvidia prioritizes hardware infrastructure dominance regardless of the underlying AI model architecture chosen by customers. - **Source:** [www.theregister.com - Articles](https://www.theregister.com) - **Published:** 2026-11-03 - **Topic:** [Data & AI](https://brifing.si/c/data-ai) - **Original article:** - **On Brifing:** https://brifing.si/s/ab93861e-437c-4ef2-9c57-c606ae178d5e ## Hardware Agnostic Strategy Nvidia maintains a neutral stance regarding the dominance of open versus closed AI models. CEO Jensen Huang emphasizes that the company benefits as long as its hardware powers the underlying infrastructure. This approach allows Nvidia to capture value from both proprietary frontier models and specialized, domain-specific open-weight alternatives simultaneously. ## Specialized Intelligence Value Huang argues that the future of artificial superintelligence lies in domain-specific adaptation rather than general intelligence alone. By utilizing open frontier models and integrating proprietary knowledge with specialized tools, companies can create highly effective systems. This strategy shifts the focus from monolithic models toward customized, high-performance AI applications. ## Fireworks.ai Infrastructure Expansion The strategic alignment between Nvidia and Fireworks.ai is underscored by a significant hardware investment. Fireworks.ai recently received 25,000 Nvidia GPUs to facilitate its expansion into European markets. This partnership supports the company's platform, which enables businesses to train and deploy models tailored to their specific operational requirements and needs. ## Rising Operational Costs Lin Qiao, CEO of Fireworks.ai, highlights that unchecked token spending is becoming a critical barrier to business scalability. Many organizations currently face unsustainable costs when relying solely on large, closed-source models. The industry is increasingly seeking more affordable alternatives to avoid the financial strain associated with high-volume AI token consumption. ## Cost-Efficient Model Performance Genspark demonstrates the potential of cost-effective AI through its Gen-1 Slides tool. By utilizing the MiniMax M3 model, the company delivers performance comparable to Anthropic’s Opus 5. This solution achieves significant efficiency, costing only $0.30 per million tokens, which represents a massive reduction compared to the $5 per million price point. ## Agent Deployment Breakthroughs Jensen Huang identifies agent harnesses as a primary breakthrough for the current year, transforming static models into active agents. He anticipates that Open Shell will play a pivotal role next year, enabling companies to deploy these agents safely and confidently. This evolution marks a shift toward more autonomous and functional AI systems. ## Robotics and Hardware Integration Looking toward future industry milestones, Huang highlights hardware-in-the-loop systems as a significant upcoming development. This focus on robotic integration aligns with Nvidia's broader investment strategy in the robotics sector. By bridging the gap between digital intelligence and physical execution, Nvidia aims to capture the next wave of industrial automation. ## Strategic Market Positioning Nvidia’s dual support for open and closed models serves as a hedge against market volatility. By providing the essential compute power for both paradigms, the company ensures its relevance regardless of which model architecture prevails. This hardware-centric strategy effectively positions Nvidia as the foundational layer for the entire global AI ecosystem. ## Sources covering this story - https://www.theregister.com/ai-and-ml/2026/11/03/nvidia-is-happy-to-take-your-money-whether-youre-using-open-or-closed-models-says-jensen-huang/5307368 — --- _Summary written by Brifing from the original reporting. Copyright in the original article belongs to its publisher; read the full piece at the link above._