JP Morgan reaffirms Zhipu overweight as AI race hits 2 trillion parameters

JP Morgan maintained its overweight rating on Zhipu (02513.HK) after Moonshot AI's Kimi K3 launch reset frontier model valuations.

"Kimi K3 resets the valuation framework for China's frontier AI models," the JP Morgan analyst said in a note dated July 22. The bank upgraded Zhipu's narrative from "domestic model leader" to "one of China's frontier AI labs."

The bank cited Zhipu's upcoming GLM-5.3 and a new flagship model with more than 2 trillion parameters as catalysts that would help the company maintain its domestic leadership. Zhipu's current GLM-5.2 remains among China's top frontier models, the note said. JP Morgan made only limited adjustments to its annual recurring revenue forecasts.

The upgrade signals institutional confidence in China's AI sector after Kimi K3 demonstrated that domestic labs can match US rivals. Moonshot's model, with 2.8 trillion parameters, overwhelmed its server infrastructure within two days of launch, forcing the company to suspend new consumer sign-ups. Zhipu faces competition from both Moonshot and Z.AI, which plans to release GLM 5.5 in August with 1 trillion parameters and a 1 million-token context window.

After recent valuation corrections, JP Morgan said the market has already priced in lower multiples and is overly pessimistic about Zhipu's commercialization trajectory. The narrative shift positions Zhipu among a select group of Chinese labs capable of competing at the global frontier, the bank said.

The maintained overweight rating suggests JP Morgan sees Zhipu well placed for the next model cycle despite intensifying competition. Investors will watch for the GLM-5.3 release and the 2T+ flagship model in the coming months as key catalysts.

This article is for informational purposes only and does not constitute investment advice.