Z.AI raises $5 billion via share sale, bond issuance as Chinese AI developers ramp up R&D efforts

Beijing-based artificial intelligence (AI) company Z.AI Co Ltd has launched a Hong Kong share placement of about $2 billion and a convertible bond sale of about $3 billion, joining other AI developers including MiniMax in raising new funds to capitalize on strong investor appetite and support their research and development (R&D) on next-generation technologies.

In a filing with the Hong Kong Exchanges and Clearing Ltd on Sunday night, Z.AI said that it is offering 21.97 million new shares at HK$714 (US$91.05) each, representing a 10 percent discount to Friday's closing price of HK$793.

The company is also selling 20.14 billion yuan ($3 billion) of zero-coupon bonds due in September 2027, the filing showed. The yuan-denominated bonds will be settled in US dollars.

The company said that by raising the funds, it aims to further broaden its shareholder equity base, seize the development opportunities presented by the evolution of general-purpose AI from model invocation towards long-horizon tasks and end-to-end task delivery, provide adequate financial support for the R&D of next-generation models, the deployment of training and inference infrastructure and the scale-up of commercialization, and ensure the group's long-term, stable and sustainable development.

Specifically, the company said that approximately 60 percent of the net proceeds will be used for the R&D of the company's next-generation GLM foundation models and its Fully Self Training system, as well as the deployment and upgrade of large-scale training and production-inference computing resources and related technology infrastructure.

In addition to Z.AI Co, another AI startup, MiniMax Group, said on July 10 that it is seeking to raise ‌HK$16.04 billion in fresh capital via a share sale and a bond issue to fund growth in its AI business.

Moonshot AI secured a $35 billion valuation after raising a larger-than-anticipated $3.5 billion in a recently closed round of financing, riding the momentum from a breakthrough Kimi K3 model that sent ripples through Silicon Valley, Bloomberg reported on July 29.

The sustained rollout of frontier models requires R&D capabilities and long-term investment to reinforce each other. Training clusters, inference infrastructure, data and talent constitute the principal cost items for large model companies. As model capabilities advance toward coding, agentic tasks and longer-horizon complex tasks, competition has extended to training data, reinforcement learning environments, inference systems and underlying infrastructure, Liu Gang, chief economist at the Chinese Institute of New Generation AI Development Strategies, told the Global Times on Monday.

Liu said China's AI developers have been focusing on enhancing intelligence density since 2025, especially the accuracy and information content of large language model outputs.