The rapid rollout highlights DeepSeek’s focus on affordable, open-weight AI systems as concerns about development speed, safety, and profitability grow across the industry.
The releases came as Anthropic CEO Dario Amodei called for a slower pace of AI development to give companies more time to address potential risks. Chinese developers, meanwhile, continued expanding their model portfolios, intensifying competition with U.S. firms such as OpenAI, Anthropic, and Google. Nikkei Asia reporting indicates that AI development cycles have shortened considerably in both markets.
China’s AI Developers Accelerate Model Releases
The 16-model tally for September reflects the expanding activity of Chinese AI laboratories, including DeepSeek, Xiaomi, and other domestic developers. According to reporting by Nikkei Asia, the count covers releases from leading companies rather than every AI model introduced in China during the month.
China’s DeepSeek and peers launch 16 AI models in a month despite Anthropic’s warning. Source: @NikkeiAsia via X
The pace of development has accelerated sharply. Nikkei’s comparison of major Chinese and U.S. AI laboratories found that the average interval between releases of upgraded models fell from 125 days between January 2023 and March 2026 to just 44 days from April through September 2026. The shorter cycle reflects growing pressure on developers to improve model capabilities and deliver new products more frequently.
A central part of China’s strategy is making AI models more accessible to developers and businesses. Open-weight models allow users to download model parameters and adapt systems to their own applications, although the degree of permitted access varies by developer and license.
Lower-cost models can help businesses deploy AI without relying exclusively on expensive proprietary services. They can also support local customization, private deployments, and specialized applications. These advantages give Chinese developers another way to compete with established U.S. providers, even as the companies continue working to improve model performance.
DeepSeek Expands Its Funding and AI Ambitions
DeepSeek’s growing financial resources are adding another dimension to China’s AI competition. Bloomberg reported in early October that the company was close to raising at least 80 billion yuan, equivalent to approximately $12 billion, in a funding round backed by investors including Tencent and battery manufacturer Contemporary Amperex Technology Co. The fundraising exceeded the company’s initial target, according to people familiar with the matter.
DeepSeek is investing tens of millions of dollars in embodied AI, expanding beyond its core business with funding from its reported $12 billion-plus financing. Source: Bloomberg via X
The financing could support further model development, computing infrastructure, and expansion into new AI applications. DeepSeek has attracted attention for its emphasis on computational efficiency and relatively inexpensive AI services, putting pressure on competitors to consider both performance and operating costs.
Its ambitions also extend beyond conventional language models. Bloomberg reported on October 12 that DeepSeek had agreed to lead a seed funding round worth tens of millions of dollars for Winverse AI, a startup developing AI systems for robots. The proposed investment would value Winverse at approximately $400 million after the funding round, although the transaction had not been finalized at the time of the report.
Founded by former Tencent scientist Hu Han, Winverse is working in the field of embodied AI, which connects artificial intelligence with physical systems that interact with the real world. The proposed investment would extend DeepSeek’s reach into robotics, where AI models must interpret their surroundings, respond to changing conditions, and perform practical tasks.
The move illustrates how Chinese AI companies are exploring opportunities beyond chatbots and conventional software. By combining model development with investments in robotics and other applications, DeepSeek could broaden the commercial uses of its technology.
AI Safety Debate Divides Development Strategies
The rapid release schedule has unfolded alongside calls from some U.S. AI executives for greater restraint. In September, Anthropic CEO Dario Amodei argued that companies should deliberately moderate the pace of improvements to advanced AI systems and strengthen safeguards against potential misuse.
“We must slow the pace at which we improve the capabilities of AI models,” Amodei wrote in an essay outlining his position. He proposed measures including independent safety evaluations, coordination among leading AI developers, and international cooperation to manage risks.
Chinese AI developers released 16 models in September, led by DeepSeek and Xiaomi, intensifying competition with U.S. firms as debate grows over AI safety, affordability, and funding. Source: @coinbureau via X
The argument centers on whether safety testing, oversight, and risk management can keep pace with increasingly capable systems. Developers face concerns that advanced AI could be misused for cyberattacks, fraud, or other harmful activities. Slower release schedules, according to advocates of greater caution, could provide additional time to identify and address these vulnerabilities.
However, the debate also takes place against the backdrop of intense commercial and geopolitical competition. Chinese developers have continued introducing models that emphasize affordability and accessibility, while U.S. companies are seeking to maintain their technological advantages.
Nikkei Asia also reported that a DeepSeek engineer questioned whether U.S. companies could be trusted to keep advanced AI systems open and affordable. The concern reflects a broader disagreement over access to powerful AI technology, particularly as companies balance commercial interests, national competition, and the interests of developers using their products.
The competing positions illustrate a difficult challenge for the industry: maintaining rapid technological progress while ensuring that new systems are adequately evaluated before deployment.
Chinese AI Valuations Rise as Funding Grows
Investment activity is reinforcing the competition. Bloomberg reported in October that Chinese AI developer Moonshot AI had reached a valuation of approximately $50 billion following its latest fundraising, with the company preparing for a potential Hong Kong initial public offering in early 2027.
DeepSeek’s reported funding round and Moonshot’s valuation reflect strong investor interest in China’s AI sector. Large financing commitments can help companies pay for computing capacity, recruit specialized researchers, and develop new products. However, the amounts being invested also raise questions about the revenue required to justify those valuations.
The commercial challenge extends to both Chinese and U.S. developers. Lower prices can attract more users and encourage broader adoption, but they can also reduce the revenue earned per customer. Companies must therefore balance competitive pricing with the substantial costs associated with computing infrastructure, research, and model training.
A faster release cycle may also require sustained spending as developers work to keep their products competitive. Whether that investment generates sufficient returns will depend on customer demand, the practical value of new capabilities, and companies’ ability to convert usage into recurring revenue.
The latest funding activity does not establish that an AI market correction is imminent. It does, however, highlight the importance of distinguishing investor expectations from demonstrated commercial performance.
U.S.-China AI Competition Faces a Commercial Test
China’s September model releases show that domestic developers are continuing to expand their AI offerings despite calls from some industry leaders for a more cautious approach. DeepSeek and Xiaomi are part of a wider group of companies seeking to compete through frequent updates, accessible models, and lower-cost services.
For U.S. technology firms, the challenge is not limited to producing more capable systems. Pricing, developer adoption, computing efficiency, and the availability of customizable models are also becoming important factors in the competition.
Chinese developers must address their own challenges, including the cost of maintaining development speed and the need to demonstrate that their products can generate sustainable returns. Greater model availability alone does not guarantee commercial adoption or long-term profitability.
The next phase of the AI race will therefore depend on more than release counts and private-market valuations. The ability to deliver useful systems, maintain appropriate safety standards, and build viable business models will help determine which companies can sustain their position in an increasingly competitive global market.
Be the first to comment