Open-Weight AI Models Surge as Enterprises Shift Away From Closed Frontier Systems

The real AI race may no longer be at the frontier

For several weeks this summer, the AI industry was fixated on Anthropic's latest frontier models and Washington's fight to control who was granted access to them. But while everyone was watching the frontier, developers kept building -- and they weren't waiting around for permission from the Anthropics and OpenAIs of the world. Chinese open-weight models accounted for 41% of downloads on Hugging Face this spring, surpassing U.S. models. On OpenRouter, the top six most popular models are all open models from Chinese firms including Tencent, Xiaomi, DeepSeek, MiniMax, and Z.ai. Anthropic's Claude Opus 4.7 trails in seventh place, at the time of this writing. And data from Vercel shows that open weight models are absorbing much of the volume-heavy infrastructure of AI apps, while closed models operate as the higher-cost, premium layer. Open models handled nearly a third of AI requests on the platform in June. Those platforms only capture one slice of the AI ecosystem; in particular, they leave out sessions hosted by major labs, which likely account for the bulk of OpenAI and Anthropic's usage. But open-source models' large and growing share of the market raises a difficult question: How much do frontier models still matter if most production AI ends up running on cheaper, customizable alternatives? Some see the growth of open-source models as a sign that the most intelligent models may end up being used for only the most specialized use cases. "Maybe in a few years, the frontier models will be for experimenting and [for] some really high value tasks, and most of the production workloads will actually be powered either by private models within companies or by open source models," Hugging Face CEO Clem Delangue said on a recent episode of Equity. Hugging Face is a platform and developer community best known for hosting, sharing, and helping companies deploy open models. Delangue says Hugging Face's customers and community members are increasingly touting the benefits of owning their own AI models rather than renting them, a trend that's picked up steam in the cold light of day after getting the bill associated with the cost of scaling closed frontier models. "If you're an AI company or a technology company, you don't want to outsource your core capabilities to another company, to a black box API that you don't control, don't have any visibility on, and don't really have any sort of ownership," Delangue said. That shift, Delangue argues, is reflected in the activity happening on Hugging Face. A new repository is created every seven seconds on the platform, which hosts almost three million public models and one million public datasets, per Delangue. That points to a different picture than the "one model to rule them all," he says. In reality, it looks more like companies using many different models, many of which are customized for their specific use case. Half of all Fortune 500 firms are using Hugging Face to deploy their own private models and open source models, he says. The growing popularity of open models coincides with a steady stream of increasingly capable releases from Chinese AI labs. Every few months, another Chinese AI company releases a powerful open-weight model that is cheaper to deploy and easier to customize than closed competitors, undercutting the economics of proprietary AI that U.S. firms have poured billions into. Most recently, Beijing-based AI company Z.ai released an open weight model called GLM-5.2 that excels at agentic coding and competes with Anthropic's latest models on identifying security vulnerabilities. Delangue isn't the only executive arguing that enterprises should avoid tying themselves to a single model provider. Microsoft CEO Satya Nadella recently warned against single provider lock-in, arguing that control of data should be a primary concern for enterprises using AI. "While the great innovation that comes from model providers having fair use rights to train models on public data is needed, I find it ironic that the status quo is to then turn around and impose restrictive terms on distillation, and to reserve the right to learn from customer usage and interaction data," Nadella said. "If learning flows in only one direction, economic value converges toward the owners of the learning infrastructure rather than the creators of the knowledge itself. Therefore, it's imperative that we distribute the learning infrastructure to every firm so that they can control their own learning loop." The rise of open models has also intensified a debate over whether increasingly capable models should be broadly available at all. Anthropic CEO Dario Amodei has argued that scaling powerful open model weights could become dangerous because once they are released, they become difficult to control. Others have argued that open models are easier to access by bad actors who could use them to spread disinformation or enact cyber or biological warfare. Delangue sees the tradeoff differently. "The biggest risk in AI is concentration of power," Delangue said. "The way you make the world safer, in my opinion, is by leveling up the playing fields and creating transparency on these models." Transparency means defenders can more easily "patch the cybersecurity risks that they already know open source models can exploit," he said. The Hugging Face executive argues that keeping powerful models closed doesn't eliminate the risks associated with advanced AI systems, in part because it's easy to get past frontier model API guardrails and to steal the weights and disseminate them openly. Restricting powerful models, Delangue argues, simply concentrates the technology in the hands of a few companies while reducing transparency into how systems work. "You don't really make it safe by keeping it behind closed doors for just a few players," Delangue said. "You make it more dangerous because you create asymmetry of power and asymmetry of capabilities."

The call for U.S. open sourced AI players

Why it matters: AI costs are ballooning, and companies don't want to be reliant on just one or two vendors for such critical tech. What we're watching: There's increasing investor interest in one "prime" solution where the U.S. would have an open-source or open-weight alternative that is as powerful as the Chinese players. * Closed models from OpenAI and Anthropic have taken the lead in the U.S. partly because of the economics. * Closed models can make more money by charging for each query, but the pathway to monetization for open models is rockier. State of play: It's still early days in the U.S. for open models. * Thinking Machines, led by former OpenAI CTO Mira Murati, dropped its first model (an open-weights one), last week. * The industry is also closely watching $25 billion Reflection AI, which has yet to release a model. Nvidia, meanwhile, has Nemotron. * OpenAI also launched open-weight models last year. What they're saying: "The open frontier is no longer just a Chinese story. It is an American one too," Benchmark's Bill Gurley wrote in a Washington Post op-ed. "Nearly everyone in the AI economy has a reason to prefer an open foundation -- everyone, that is, except the big incumbents Anthropic and OpenAI, whose fortunes depend on keeping it closed." Between the lines: Developments in new AI models are happening faster than ever. * The conversation today has centered on Kimi K3, but the leading model in three to six months could be a whole different story. * "The most important things to understand about the new AI race right now is that every metric is changing really fast right now," Artificial Analysis CEO Micah Hill-Smith tells Axios. Costs are going down inside of closed-source AI labs, while services companies are emerging aiming to also bring down bills. The bottom line: Volatility is a big part of the AI industry right now. We're all just getting used to it.

Open-weight AI models drive shift to data control

Together AI positions open-weight AI models as the enterprise moat for cost, control and IP Enterprises racing to deploy AI at scale are discovering that the biggest constraint isn't model capability anymore -- it's control. As agentic AI moves from experimentation into core business processes, companies are rethinking whether handing proprietary data to closed frontier models is a risk worth taking, opening the door for open-weight AI models. That shift is fueling explosive growth for the companies building the infrastructure layer beneath open-source AI. Token usage on open-weight models has surged as enterprises weigh cost, compliance and intellectual property against the convenience of closed systems, according to Vipul Ved Prakash (pictured), co-founder and chief executive officer of Together AI Inc., which recently raised $800 million in Series C funding at an $8.3 billion valuation. "One of the things that we have seen over the last year is there's been almost a stampede towards open-weights models, which we serve and we allow our customers to post-train and adapt to their data," Prakash said. "We've seen a 10,000-times increase in the number of tokens being processed through open-source models. I think they have really become now a workhorse of agentic AI in a way that was just not there a year ago." Prakash spoke with theCUBE's John Furrier at the RAISE Summit in Paris, during an exclusive broadcast on theCUBE, SiliconANGLE Media's livestreaming studio. They discussed the rise of open-weight AI models, enterprise agent harnesses and how sovereignty concerns are reshaping AI infrastructure decisions. (* Disclosure below.) Open-weight AI models reshape enterprise cost and control equations Cost is a major driver of the shift, but it's not the only one. Together AI's customers see cost differences between open and closed models ranging from six to 60 times, Prakash said, a gap that becomes decisive once AI runs at production scale rather than in a demo. "[Open-weight models] are important for a couple of reasons," Prakash said. "One is cost. ... The other is control. These models can be run in the compute environment that the customer wants, following the compliance and data loss and the security requirements for the customer." Enterprises increasingly worry that sending proprietary business processes into closed frontier models effectively hands competitors a blueprint, Prakash noted, pointing to public comments from Palantir Technologies Inc. CEO Alex Karp on the same tension. That anxiety is growing as Together AI's own volume signals how fast agentic workloads are scaling. "We were serving 30 billion tokens a month 9 months ago," he said. "We are serving over 400 trillion tokens a month now. So, there is an incredible appetite. It's become a compute-bound business." Enterprises are responding by building their own "harnesses" -- orchestration loops that let them swap models underneath an application with near-zero switching cost, Prakash explained. That flexibility, paired with data control, is turning open infrastructure into a durable competitive advantage rather than just a budget line item. "You are not sharing your data with a company that trains models," Prakash said. "You have complete control on data residency, what happens with that data, and you can still mix and match multiple models within your harnesses to get the best results. I think this starts becoming a moat in that you're deploying AI effectively in the enterprise ... all the while you're also creating these AI assets that you now own, and it becomes part of your intellectual property." Here's the complete video interview, part of SiliconANGLE's and theCUBE's coverage of RAISE Summit: (* Disclosure: TheCUBE is a paid media partner for the RAISE Summit event. Neither Solidigm, the headline sponsor of theCUBE's event coverage, nor other sponsors have editorial control over content on theCUBE or SiliconANGLE.)