\n Yet despite the fanfare, some leading Chinese AI developers have warned that China may have fallen further behind in developing frontier models. Experts point to restricted access to advanced chips and limited capital as lingering constraints.\n <\/p>\n
\n Lin was not alone. Tang Jie, founder of one of the prominent Chinese AI startups Z.ai, also known as Zhipu, said the performance gap between Chinese and US models \u201cmay be widening.\u201d\n <\/p>\n
\n \u201cIn some areas we may be doing fairly well, but we also need to acknowledge the challenges and gaps we still face,\u201d he said at the same Beijing meeting.\n <\/p>\n
\n But that assessment does not mean China\u2019s AI industry is stagnating.\n <\/p>\n
\n Constraints in accessing high-performance chips and capital as well as the country\u2019s unique tech ecosystem have fueled a divergent strategy from the US \u2013 making AI models available for public use, or open source.\n <\/p>\n
\n The strategy, which Beijing and developers see as a way to accelerate progress and compete with US rivals, has helped Chinese companies make notable gains. Companies are aggressively rolling out AI applications based on these models for real world uses. Industries are integrating the technology in manufacturing, e-commerce and robotics.\n <\/p>\n
![\"'Minions'](\"https:\/\/www.newsbeep.com\/au\/wp-content\/uploads\/2026\/02\/gettyimages-1988940918.jpg\")
<\/p>\n
\n In a televised new year address<\/a>, Chinese leader Xi Jinping hailed the country\u2019s fast rising innovative capabilities, citing AI models that are \u201cracing ahead\u201d and what he called \u201cbreakthroughs\u201d in homegrown chips, as Beijing pushes for technological self-reliance.\n <\/p>\n \n Qwen, for example, overtook Meta\u2019s Llama last September as the most-downloaded open model on Hugging Face, a major platform for AI models and tools. Even American firms like Airbnb have embraced it to power AI customer service.\n <\/p>\n \n An unprecedented wave of Chinese AI listings has also unfolded in Hong Kong. In January, unicorn startups Z.ai and MiniMax, both developers of competitive open models, went public, raising $560 million and $620 million respectively, with their share prices surging.\n <\/p>\n \n Global tech companies are taking notice. In December, Meta announced<\/a> it would acquire Manus, an AI agent firm founded in China and later relocated to Singapore. While the deal has triggered Beijing\u2019s regulatory review<\/a> process \u2013 raising the possibility it could be unwound, it is a testament to the advances in Chinese AI technology.\n <\/p>\n \n DeepSeek, which emerged as a poster child for China\u2019s AI boom after releasing an open model a year ago that delivered near industry-leading performance with far fewer resources, is also reportedly set to unveil a new model with enhanced coding capabilities later this month.\n <\/p>\n \n A key driver of China\u2019s AI momentum stems from its aggressive embrace of open models, a shift sparked by DeepSeek\u2019s breakout success. Since then, Chinese AI labs have jumped on the open-model bandwagon, in contrast to their US rivals who have largely eschewed that approach in order to protect their intellectual property.\n <\/p>\n \n The turn to open models in China \u201chas dramatically reduced costs for developers and enterprises,\u201d said Poe Zhao, analyst and founder of Hello China Tech, a newsletter focused on the country\u2019s tech industry.\n <\/p>\n \n \u201cCloud providers like Alibaba use open models to drive cloud adoption, startups use openness to build developer ecosystems quickly.\u201d\n <\/p>\n \n The boom of open models in China prompted their global usage to skyrocket from a miniscule 1.2% in late 2024 to as much as nearly 30% last year, according to a late 2025 study<\/a> by OpenRouter, a marketplace for AI models.\n <\/p>\n \n Alibaba alone, for example, has released more than 400 open Qwen models, yielding over one billion downloads by earlier this month, according to the company.\n <\/p>\n \n \u201cOpen sourcing has to a certain extent become an industry consensus in China,\u201d said Lian Jye Su, chief analyst at Omdia, a technology research firm.\n <\/p>\n \n The strategy not only reduces costs but also lessens Chinese companies\u2019 exposure to geopolitical risks.\n <\/p>\n \n \u201cIf one day the entire company were to be sanctioned (by the US), then at least the products based on its open-sourced models could still be used by others,\u201d Su said.\n <\/p>\n \n There is also a hard economic reality behind the strategy.\n <\/p>\n \n \u201cThere\u2019s no other choice. The reason is that Chinese people, like consumers, as well as enterprises, do not pay for software,\u201d said Jenny Xiao, a partner at Leonis Capital, an AI-focused venture capital firm.\n <\/p>\n \n But while Chinese models reign supreme in the open model space, closed models developed by US heavyweights \u2013 including OpenAI\u2019s GPT, Google\u2019s Gemini and Anthropic\u2019s Claude \u2013 continue to dominate overall performance benchmarks. Closed models still account for roughly 70% of total downloads, according to OpenRouter.\n <\/p>\n Bottlenecks remain<\/p>\n \n The performance gap between Chinese and US models, while not enormous, persists largely due to constraints on computing power and capital.\n <\/p>\n \n \u201cOpenAI, Anthropic and other American companies are putting massive compute into next-gen research, whereas we are relatively stretched thin,\u201d Alibaba\u2019s Lin said.\n <\/p>\n![\"Signage](\"https:\/\/www.newsbeep.com\/au\/wp-content\/uploads\/2026\/02\/gettyimages-2239628142.jpg\")
<\/p>\n
![\"A](\"https:\/\/www.newsbeep.com\/au\/wp-content\/uploads\/2026\/02\/ap25324459292175.jpg\")
<\/p>\n![\"Robotic](\"https:\/\/www.newsbeep.com\/au\/wp-content\/uploads\/2026\/02\/gettyimages-2251495931.jpg\")
<\/p>\n