Guozhen AIGlobal AI field notes and model intelligence

Realtime AI News

Zhipu AI Stock Rises After Lock-Up Expiry as IPO Valuation Question Looms

Zhipu AI's stock opened low but recovered on its lock-up expiration day, while market analysts flag its ultra-high valuation as a potential hurdle for a planned A-share IPO. The company, dubbed a "12x bull stock," is one of China's most prominent large-language-model startups.

Published
智谱解禁日股价低开高走,超高估值或成A股IPO挑战
Image source: zhipuai.cn

Zhipu AI, one of China's leading large language model companies, saw its stock open lower but recover through the trading day on its lock-up expiration date, drawing market attention to the company's IPO prospects. Since its listing, Zhipu's share price has surged over 12-fold, earning it the nickname "12x bull stock" among Chinese investors.

Lock-up expiration days typically allow early investors and insiders to sell their shares on the open market, often creating downward price pressure. Despite opening lower, Zhipu's stock recovered during the session, suggesting sustained investor confidence in the company's long-term value.

However, market analysts point out that Zhipu's sky-high valuation could become a significant obstacle in its A-share IPO process. As an AI model company that has not yet achieved large-scale profitability, Zhipu's valuation already far exceeds many traditional tech enterprises, which may face challenges under mainland China's capital market pricing frameworks.

Zhipu AI is one of China's "Six Little Tigers" of large language models, incubated by a Tsinghua University team. It has developed the GLM series of models and completed multiple funding rounds backed by internet giants including Meituan, Alibaba, and Tencent, as well as state-owned capital.

The valuation of AI model companies remains a global challenge. While overseas peers like OpenAI and Anthropic continue to see their private-market valuations climb, most have yet to establish stable, large-scale revenue models — creating a persistent gap between high valuations and profitability reality.

Key questions ahead: Can Zhipu successfully navigate its A-share IPO process? Will regulators and secondary-market investors accept its current valuation level? And when will China's large model companies prove their business models are sustainable?

Why it matters

Zhipu's lock-up expiration performance and IPO valuation debate reflect capital market uncertainty around pricing Chinese AI model companies. As one of the "Six Little Tigers," Zhipu's listing path will serve as a critical reference for the funding trajectory of China's AI startups.

智谱ZhipuA股IPOLock-up
Back to AI Daily

Nearby Updates

All

07/09, 09:45

StepStar Unveils Next-Gen AI Agent Terminal, Sources Say It Includes Both Software and Phone Hardware

Chinese AI company StepStar (阶跃星辰) has officially announced a next-generation AI agent terminal product, with sources reporting the project includes both an AI agent software system and custom smartphone hardware. The move places StepStar among a growing list of Chinese tech companies racing to build AI agent-powered devices.

07/09, 09:54

GPT-Live Demonstrates Real-Time Simultaneous Interpretation, Viral Video Stuns Viewers

GPT-Live has showcased real-time simultaneous interpretation capabilities, translating speech across languages with minimal latency during live conversations. A viral video featuring an elderly woman spontaneously debating with the AI has amazed viewers across the internet.

07/09, 10:32

Quantitative Pai Releases Physical World Foundation Model, from Restaurant Kitchens to General Physical AI

Quantitative Pai has officially launched a physical world foundation model capable of applications ranging from restaurant kitchens to general physical AI scenarios. The company plans to commercialize the model by selling its capabilities to enterprise clients.

07/09, 10:53

Study Reveals Public GitHub Issues Can Trick AI Into Leaking Private Data, Bypassed With a Single Word

Security researchers have discovered that public GitHub Issues can be weaponized to trick AI models into leaking private data. Attackers need only a single carefully chosen keyword to bypass existing safety guardrails, raising new concerns about AI data security.