21ьз╕ C

Zhipu, Jiangmen, QuantGroup and Other Investors Join Forces to Back RMB 1 Billion Cognitive Intelligence Venture Capital Fund for Physical AI Ecosystem Development

News Provided by ACN Newswire2026-10-22

Hong Kong, Oct 22, 2026 - (ACN Newswire) -Competition in AI has evolved from contests over individual technologies into a new phase defined by coordination across foundational ecosystems and industry value chains. Zhipu, a leading large-model developer, has again expanded its footprint as a limited partner (LP) in the primary market by teaming up with Jiangmen Ventures to establish a cognitive intelligence venture capital fund. QuantGroup (02685.HK), a provider of world foundation models for Physical AI, is among the fund's LPs.

Corporate registration records show that Zhejiang Zhipu Jiangmen Cognitive Intelligence Venture Capital Partnership (Limited Partnership) was incorporated in Shangcheng District, Hangzhou, on 24 September with total subscribed capital of RMB 1 billion. Its business scope covers venture capital, limited to investments in unlisted companies. The fund combines local state-owned capital, industrial investors and specialist investment firms. Two local government-backed limited partners, Hangzhou Shangcheng Linghang Venture Capital Co., Ltd. and Hangzhou Urban Investment Industry Development Investment Partnership (Limited Partnership), hold a combined 60% interest. The other partners are Zhejiang Zhipu Xinpian Technology Co., Ltd., an affiliate of Zhipu; Hainan QuantGroup Technology Co., Ltd., an affiliate of Hong Kong-listed QuantGroup (02685.HK); Hainan Hezun Investment Co., Ltd.; and Hangzhou Jizhi Enterprise Management Consulting Co., Ltd. The latter also serves as the fund's executive partner, and its appointed representative is Du Feng, a founding partner of Jiangmen Ventures. The deadline for partners to pay in their subscribed capital is September 2031.

Industry observers generally expect that collaboration among leading companies raises the success rate of ecosystem-building, reinforcing a Matthew effect in which the strong become stronger. Companies that are first to form ecosystem alliances can secure early access to downstream use cases and early-stage projects. Zhipu has pursued precisely this approach. In 2025, it partnered with Unilumin Group (300232.SZ), a leader in optoelectronic smart display technology, and MetaView to establish Shenzhen Zhixian Robot Technology Co., Ltd., building an innovation ecosystem for AI-powered smart terminals. In the energy sector, iFlytek worked with PetroChina, China Mobile and Huawei to develop the Kunlun large model, helping PetroChina set a new benchmark for AI in energy.

QuantGroup's presence among the LPs has attracted considerable attention. As Hong Kong's first listed "intelligent species" company, QuantGroup positions itself as a provider of world foundation models for Physical AI. It has pursued a two-pillar strategy combining AI-enabled e-commerce solutions with Physical AI, using cash flow from its established consumer business to fund frontier R&D. For infrastructure providers, ecosystem building is a core capability: connecting upstream model developers, midstream embodied-robotics developers, downstream use cases and capital providers to secure a central position in the Physical AI ecosystem during its early development. At the ecosystem level, QuantGroup works with partners including ByteDance's Volcengine, Beijing Ruihong Embodied Intelligence Robot Technology Co., Ltd., an affiliate of Tungkong Inc. (002117.SZ), and Chewujie to build an embodied AI ecosystem. This has extended its commercial reach from restaurant kitchens to brick-and-mortar venues such as car showrooms. Its role as an LP in the cognitive intelligence fund established by Zhipu and Jiangmen Ventures marks another step toward moving upstream in the Physical AI value chain and gaining earlier access to large-model capabilities and early-stage projects.

Building an ecosystem with the first publicly listed large-model company naturally improves the odds of success, as the two sides can combine their strengths, complement one another and develop benchmark industry solutions quickly. Zhipu (02513.HK) listed on the Hong Kong Stock Exchange on 8 January, becoming the world's first publicly listed large-model company. In May, Beijing Zhipu Future Technology Co., Ltd., a wholly owned subsidiary of Zhipu, and Guotai Junan Innovation Investment Co., Ltd., a wholly owned subsidiary of Guotai Haitong, jointly established an RMB 540 million AI industry fund in Zhangjiang, Shanghai. Earlier, in 2024, Zhipu led the establishment of the Z Fund (LinkX Capital), an AGI ecosystem fund with RMB 1.5 billion of assets under management at its first close. According to Zhipu's 2025 financial report, LinkX Capital (Z Fund) had an aggregate value of RMB 1.3027 billion at the end of 2025 and had invested in more than 30 AI companies.

Jiangmen Ventures, which is partnering with Zhipu on the new fund, is an early-stage venture capital firm focused on discovering technology-driven start-ups. It primarily invests in foundational AI technologies and their industrial applications, and was one of Zhipu's Series A investors. The two organizations have a long-standing relationship. The "cognitive intelligence" in the fund's name aligns with Zhipu's large-model technology roadmap. The fund's location in Hangzhou also builds on Zhipu's ties with the city. In March 2025, Zhipu raised more than RMB 1 billion in strategic financing, with participation from Hangzhou municipal state-owned investors including Hangzhou City Investment Industrial Fund and Shangcheng Capital.

Venture capital professionals note that, for companies expanding into Physical AI, investing as an LP in an industry fund led by a leading large-model company can generate financial returns and provide early access to upstream model capabilities and early-stage projects. As large-model competition shifts from pure performance to ecosystem strength, more industrial companies are entering the primary market as investors.

U.S. National Debt Clock

national debt