Tang Jie, CEO of China’s leading AI startup Zhipu is quite clear that China is not months behind America in the race for frontier AI models. The prediction has been made in what can be termed as a rare social media exchange between Zhipu founder and CEO Tang Jie and Tesla CEO Elon Musk. In a recent interaction on X, formerly Twitter, Elon Musk wrote in reply to a post on China’s timeline to launch an AI model to rival Anthropic’s new Fable 5: “Probably Q1”. To this, Jie Tang replied: “won’t take that long.”Tang Jie wrote that the Chinese rival to Anthropic’s Fable 5-class AI model is closer than what the world’s richest man Musk thinks. To this Musk replied, arguing that even if Chinese models catch up to Fable 5 in benchmarks, “true usefulness” is a different matter. He stressed that Anthropic’s focus on “maximizing useful intelligence” “shows up in revenue.” To this, Tang Jie replied: “Focus is all we need, especially focusing on what intelligence truly is.” The exchange sparked a fresh debate over the status of artificial intelligence (AI) gap between the two countries — US and China. Or rather basically how far (or behind) China is vs America in the AI race. Curbs on Anthropic announced by the Trump admin have also revived debates around AI talent race between the U.S. and China.As for who Jie Tang is? He is the founder of Beijing-based Z.ai, formerly known as Zhipu AI, that is one of the leading AI startups in China. The company recently released its latest AI model, GLM-5.2. According to the company’s benchmarks, GLM-5.2 has almost the same performance as Anthropic’s Opus 4.7 to 4.8, which launched in April and May of this year, respectively. Zhipu’s GLM-5.2 has also reportedly consistently outperformed OpenAI’s GPT-5.5. Zhipu’s GLM-5.2 is its latest and most capable open-source large model, with no usage restrictions. Announcing the launch, the company said, “Cutting-edge intelligence should not belong to only a few, nor should it be withdrawn at any time,” the company said. “It should be open, available, extensible and built to serve every developer.”
Chinese AI model turning heads in America
A 744-billion-parameter model, Zhipu’s GLM-5.2 model ranked world No. 2, behind Fable 5, in the front-end coding rankings of benchmark provider “CodeArena.” With this release, Zhipu is said to have become the world’s third-ranked lab by Artificial Analysis Intelligence Index (AAII), following Anthropic and OpenAI. Jefferies called GLM-5.2 “a milestone for Chinese AI.” It marks the first time that a Chinese model has reached the global top three.The exchange between Elon Musk and Zhipu CEO also moved markets, with Zhipu’s stock, traded in Hong Kong, surged as much as 42% intraday that day. The company’s market capitalization surpassed HK$1 trillion (about 196 trillion won). Its cumulative gain since the January listing exceeds 1,700%. In a research note last week, JPMorgan projected a more than 534 per cent revenue surge for the AI firm this year and predicted it would swing to a profit by 2028, reversing its previous forecast of a net loss.Shares of Zhipu are up 170% since the end of March in Hong Kong. Boosted by China’s self-reliance push on AI, the company ranks among the top three performers on the Hang Seng Tech Index this year, despite lack of profits. Zhipu is also the faraway leader with its gain of more than 1,500% on the Stock Index.As for Claude Fable 5, Anthropic launched its latest AI models — Fable 5 and Mythos 5 — in early June. However, three days after the release of Fable 5, the US government put an export control directive on the model and barred all foreign nationals, including Anthropic’s own employees, from accessing it. Since Anthropic determined that it cannot guarantee total compliance with the order, it decided to pull both Fable 5 and Mythos 5 from across the globe.
Leave a Reply