Chinese foundation model company Zhipu AI set its Hong Kong IPO price at HK$116.2 per H-share, according to a Dec. 30, 2025 global offering announcement. The 37,419,500-share offer could raise about HK$4.35 billion, with trading on the HKEX main board scheduled to start on January 8, 2026.
This article aggregates reporting from 1 news source. The TL;DR is AI-generated from original reporting. Race to AGI's analysis provides editorial context on implications for AGI development.
Zhipu AI’s decision to list in Hong Kong at a price implying roughly US$559 million in new capital is a milestone for China’s homegrown foundation model ecosystem. ([tech.sina.cn](https://tech.sina.cn/2025-12-30/detail-inhepszs7244336.d.html?vt=4&wm=1184%3Fp)) Unlike earlier AI listings built around applications or cloud services, this float is explicitly about funding large-scale model research and commercialization for a domestic LLM champion. It’s a bet that there is enough investor appetite—inside and outside mainland China—for a pure-play model builder positioned as an alternative to OpenAI, Google, and DeepSeek.
Strategically, the raise gives Zhipu more runway to train larger multimodal models, expand inference infrastructure, and deepen vertical solutions in areas like finance, government and manufacturing. In a world of tightening US export controls and fragmented AI supply chains, having a well‑capitalized Chinese model provider with access to Hong Kong capital markets is important for both national AI ambitions and regional investors seeking exposure to non-US foundational models.
For the broader race to AGI, an IPO-backed Zhipu validates that public equity markets are now willing to underwrite multi-hundred-million-dollar model development programs, not just at US hyperscalers but in China as well. That competition could accelerate architectural innovation and scaling experiments, particularly if Zhipu channels capital into more efficient training and alignment techniques that don’t rely on US GPUs.

