Qunhe Technology (群核科技) surges on Hong Kong debut as it pivots from design software to spatial intelligence

Trading debut and fundraising

Qunhe Technology (群核科技) (stock code: 00068) made a striking entry on the Hong Kong Stock Exchange, opening more than 170% up on its first day of trading and pushing market value past HK$35.1 billion. The company issued about 161 million shares in the global offering, with a top price of HK$7.62 per share and a maximum expected raise of HK$1.227 billion; demand in the public tranche was reported at some 1,590.56 times oversubscription. It has been reported that Qunhe is the first of the informal “Hangzhou Six Little Dragons” cohort of frontier tech startups to reach IPO — a local label that highlights the city’s growing startup cluster.

Financials and market position

Qunhe’s recent results show a recovery: revenue climbed from RMB 664 million in 2023 to RMB 820 million in 2025, gross margin reached 82.2%, and adjusted net profit swung to RMB 57.13 million after a loss the prior year. The company’s core home‑design software “Kujiale” (酷家乐) gave it a 23.2% share of the China space‑design software market by 2024 revenue and remains the bulk of its business — subscription services accounted for 96.9% of 2025 revenue, while spatial‑intelligence products contributed only 3.1% and thus remain nascent.

Strategy, R&D and geopolitical context

Qunhe is betting heavily on R&D and infrastructure to extend beyond software into spatial intelligence. The firm says it has invested more than RMB 1 billion over three years, maintaining a 37.5% R&D expense ratio in 2025 and building a library of more than 479 million 3D models and design elements. Chairman Huang Xiaohuang has endorsed an in‑house GPU cluster strategy to control compute costs — but why build internal GPU farms now? Part of the answer is geopolitical: amid US export controls and broader technology tensions that constrain access to cutting‑edge chips, Chinese AI and cloud firms increasingly pursue self‑reliant compute stacks. Qunhe has also open‑sourced SpatialLM and launched a SpatialGen model as it seeks to translate its 3D data advantage into generative spatial intelligence — a long game that investors will be watching closely.