A powerful newopen-sourceartificial intelligence model created by Chinese startup DeepSeek has shaken Silicon Valley over the past few days. Packed with cutting-edge capabilities and developed on a seemingly tiny budget, DeepSeek’s R1 is prompting talk of an impending upheaval in the tech industry.

To some people, DeepSeek’s rise signals that the US has lost its edge in AI. But a number of experts, including executives at companies that build and customize some of the world’s most powerful frontier AI models, say it's a sign of a different kind of technological transition underway.

Instead of trying to create larger and larger models that require increasingly exorbitant amounts of computing resources, AI companies are now focusing more on developing advanced capabilities, like reasoning. That has created an opening for smaller, innovative startups such as DeepSeek that haven’t received billions of dollars in outside investment. “It’s a paradigm shift towards reasoning, and that will be much more democratized,” says Ali Ghodsi, CEO of Databricks, a company that specializes in building and hosting custom AI models.

“It’s been clear for some time now that innovating and creating greater efficiencies—rather than just throwing unlimited compute at the problem—will spur the next round of technology breakthroughs,” says Nick Frosst, a cofounder of Cohere, a startup that builds frontier AI models. “This is a clarifying moment when people are realizing what's long been obvious.”

Thousands of developers and AI enthusiasts flocked to DeepSeek’s website and its official app in recent days to try out the company’s latest model and shared examples of its sophisticated capabilities on social media. Shares in US tech firms, including the chipmaker Nvidia, fell in response on Monday as investors began to question the vast sums being poured into AI development.

DeepSeek’s technology was developed by a relatively small research lab in China that sprang out of one of the country’s best-performing quantitative hedge funds. A research paper posted online last December claims that its earlier DeepSeek-V3 large language model cost only $5.6 million to build, a fraction of the amount its competitors needed for similar projects. OpenAI has previously said that some of its models cost upwards of $100 million each. The latest models from OpenAI as well as Google, Anthropic, and Meta likely cost considerably more.

The performance and efficiency of DeepSeek’s models has already prompted talk of cost cutting at some big tech firms. One engineer at Meta, who asked not to be named because they were not authorized to speak publicly, says the tech giant will most likely try to examine DeepSeek’s techniques to find ways to reduce its own expenditure on AI. “We believe open source models are driving a significant shift in the industry, and that’s going to bring the benefits of AI to everyone faster,” a spokesperson for Meta said in a statement. “We want the US to continue to be the leader in open source AI, not China, which is why Meta is developing open source AI with our Llama models which have been downloaded over 800 million times.”

The true price of developing DeepSeek’s new models remains unknown, however, since one figure quoted in a single research paper may not capture the full picture of its costs. “I don't believe it's $6 million, but even if it's $60 million, it's a game changer,” says Umesh Padval, managing director of Thomvest Ventures, a company that has invested in Cohere and other AI firms. “It will put pressure on the profitability of companies which are focused on consumer AI.”

Shortly after DeepSeek revealed the details of its latest model, Ghodsi of Databricks says customers began asking whether they could use it as well as DeepSeek’s underlying techniques to cut costs at their own organizations. He adds that one approach employed by DeepSeek’s engineers, known as distillation, which involves using the output from one large language model to train another model, is relatively cheap and straightforward.

​Padval says that the existence of models like DeepSeek's will ultimately benefit companies looking to spend less on AI, but he says that many firms may have reservations about relying on a Chinese model for sensitive tasks. So far, at least one prominent AI firm, Perplexity, has publicly announced it's using DeepSeek’s R1 model, but it says it is being hosted “completely independent of China.”

Amjad Massad, the CEO of Replit, a startup that provides AI coding tools, told WIRED that he thinks DeepSeek’s latest models are impressive. While he still finds Anthropic’s Sonnet model is better at many computer engineering tasks, he has found that R1 is especially good at turning text commands into code that can be executed on a computer. “We’re exploring using it especially for agent reasoning,” he adds.

DeepSeek’s latest two offerings—DeepSeek R1 and DeepSeek R1-Zero—are capable of the same kind of simulated reasoning as the most advanced systems from OpenAI and Google. They all work by breaking problems into constituent parts in order to tackle them more effectively, a process that requires a considerable amount of additional training to ensure that the AI reliably reaches the correct answer.

A paper posted by DeepSeek researchers last week outlines the approach the company used to create its R1 models, which it claims perform on some benchmarks about as well as OpenAI’s groundbreaking reasoning model known as o1. The tactics DeepSeek used include a more automated method for learning how to problem-solve correctly as well as a strategy for transferring skills from larger models to smaller ones.

One of the hottest topics of speculation about DeepSeek is the hardware it might have used. The question is especially noteworthy because the US government has introduced a series of export controls and other trade restrictions over the last few years aimed at limiting China’s ability to acquire and manufacture cutting-edge chips that are needed for building advanced AI.

In a research paper from August 2024, DeepSeek indicated that it has access to a cluster of 10,000 Nvidia A100 chips, which were placed under US restrictions announced in October 2022. In a separate paper from June of that year, DeepSeek stated that an earlier model it created called DeepSeek-V2 was developed using clusters of Nvidia H800 computer chips, a less capable component developed by Nvidia to comply with US export controls.

A source at one AI company that trains large AI models, who asked to be anonymous to protect their professional relationships, estimates that DeepSeek likely used around 50,000 Nvidia chips to build its technology.

Nvidia declined to comment directly on which of its chips DeepSeek may have relied on. “DeepSeek is an excellent AI advancement,” a spokesman for Nvidia said in a statement, adding that the startup's reasoning approach “requires significant numbers of Nvidia GPUs and high-performance networking.”

However DeepSeek’s models were built, they appear to show that a less closed approach to developing AI is gaining momentum. In December, Clem Delangue, the CEO of HuggingFace, a platform that hosts artificial intelligence models, predicted that a Chinese company would take the lead in AI because of the speed of innovation happening in open source models, which China has largely embraced. “This went faster than I thought,” he says.