Chinese technology firms are positioning their latest artificial-intelligence systems as direct rivals to leading American models, even as US export restrictions keep the top Western systems out of the Chinese market.
According to a TechCrunch report by Kate Park, two Chinese efforts are now claiming parity with Anthropic's banned Mythos and Fable 5 models. On Wednesday, the Chinese cybersecurity firm 360 reportedly unveiled a tool called Tulongfeng, which the company says can go head-to-head with Anthropic's systems. Sakana AI's model, Fugu, is making a similar claim. Both pitches arrive against the backdrop of a US export ban that limits Chinese access to the most advanced American AI.
The theme extends beyond a single launch. Coverage from Yahoo Finance and BeInCrypto carries the same headline message: China's AI models are gaining ground on both Anthropic and OpenAI, the two firms that have set much of the pace in the field.
The competition is not limited to chatbots and security tools. According to Memeburn, a Chinese system referred to as LineShine has beaten a US supercomputer, though the outlet notes that the broader AI race looks different heading into 2026.
Taken together, the reports point to a narrowing gap between Chinese and American AI despite Washington's attempts to slow China's progress through trade controls. It is worth noting these are largely claims made by the Chinese firms themselves, and the sources do not provide independent benchmark results confirming them.
Why it matters: if Chinese developers can credibly match top US models while those models are barred from China, export restrictions may be accelerating, rather than preventing, the rise of homegrown competitors.