Zhipu AI's market value crossed $128 billion on Monday after its open-source model ranked second worldwide on a closely watched coding benchmark, trailing only Anthropic's top system, Fable 5.
Key Points:
- Zhipu AI's market cap topped $128 billion as its GLM-5.2 model placed second on the Code Arena coding benchmark.
- The open-weight model sits behind Anthropic's Claude Fable 5 and within one point of Claude Opus 4.8 on long coding tasks.
- Founder Tang Jie said a Chinese rival to Fable 5 will arrive sooner than Elon Musk forecast.
Zhipu Stock Surges 42% On GLM-5.2
The Beijing developer's Hong Kong-listed shares jumped as much as 42% on Monday, lifting its capitalization past 1 trillion Hong Kong dollars, or about $128 billion. The stock has now climbed more than 800% since its January listing. Investors read the benchmark result as concrete backing for the run, after one major bank lifted its target and dubbed the stock an AI winner, while another opened coverage with a buy rating.
The rally followed the recent debut of GLM-5.2, a 744-billion-parameter mixture-of-experts model built for coding jobs that run for hours. It launched Jun. 13 under a permissive MIT license, pairing a 1-million-token context window with free weights that anyone can download, fine-tune and run with no regional limits.
On Code Arena's front-end coding board, the open-weight model placed second globally, behind only Anthropic's Claude Fable 5. It trails Claude Opus 4.8 by roughly one point on the hardest agentic tests, yet beats GPT-5.5 on real-world bug fixing and nails a near-perfect 99.2% on a flagship math exam. Subscriptions start near $10 a month, about a tenth of comparable Western pricing, and per-token API rates undercut Opus by a similar margin.
Also Read: Strategy's Bitcoin Machine Keeps Running As Treasury Nears 850K BTC
Tang Jie Disputes Musk On AI Gap
The ranking drew a rare public exchange. Founder Tang Jie told Elon Musk that a Chinese rival to Fable 5 won't take that long. Musk had estimated such a model would probably surface in the first quarter of next year.
Analysts read the moment as a sign of a fast-closing divide. Stanford's 2026 AI Index pegged the gap between the best American and Chinese systems at 2.7 percentage points, though the lead holds wider on the toughest reasoning tasks.
Anthropic Export Ban Reset The Race
Washington's pressure on Chinese AI has been building for months. Lawmakers opened a House inquiry in May into PRC-origin models deployed near critical infrastructure, naming Zhipu among the firms under review. Cloud users also face China's intelligence law, which can compel access to the data they route through the service.
The current standoff took shape just before GLM-5.2 arrived. On Jun. 12, the US Commerce Department ordered Anthropic to cut foreign access to its Fable 5 and Mythos 5 models within 48 hours, citing a narrow security concern. Both went dark worldwide for every customer, foreign and domestic alike, and Zhipu released its open model the next day.
Read Next: Ethereum Bears Load Up $85M In Shorts As Traders Chase A Rebound