Zhipu raised $5 billion through a zero-interest-rate and premium-converted-equity swap, primarily for computing infrastructure and model research and development.
On September 13, Zhipu announced that it had completed a new round of financing of approximately US$5 billion (approximately RMB 33.5 billion), consisting of approximately US$2 billion in share placement and approximately US$3 billion in zero-coupon convertible bonds. This comes just two months after its previous share placement of approximately US$4 billion in July.
This marks Zhipu's third equity financing move since its listing on the Hong Kong Stock Exchange in January of this year. From its IPO to two subsequent rounds of financing, the pace of fundraising has been accelerating, and the flow of funds has been highly consistent—computing power, model development, and infrastructure.

Financing structure: Discounted placement plus zero-coupon premium convertible bonds
This financing will be conducted in two independent parts, which are not conditional on each other.
Regarding the share placement , Zhipu will place up to 21.965 million new H shares at HK$714 per share to no fewer than six institutional investors. The placement price represents a discount of approximately 9.96% to the closing price of HK$793 on the previous trading day (September 11) and a discount of approximately 19.95% to the average price of HK$891.90 over the past five trading days . The placed shares represent approximately 4.50% of the enlarged issued share capital, and the net proceeds are expected to be approximately HK$15.664 billion.

Regarding convertible bonds , Zhipu issued zero-coupon convertible bonds with a total principal amount of RMB 20.14 billion, maturing in September 2027, at an issue price of 100.5% of the principal , redeemable at principal upon maturity. The initial conversion price is set at HKD 892.50 per share, representing a premium of approximately 12.55% over the closing price before the announcement. Assuming full conversion, the bonds will be convertible into approximately 26.365 million new H shares, representing approximately 5.36% of the enlarged issued share capital. The net proceeds from the bonds are expected to be approximately US$3.011 billion.


Zero-interest means that holders do not receive interest, while premium conversion means that conversion is only economically meaningful if the share price rises further—this structure conveys expectations for the long-term share price to the market.
Where is the money flowing: 60% bet on computing power and next-generation models
According to the announcement, Zhipu intends to distribute the net proceeds in the following proportions:
- Approximately 60% is used for the research and development of the next-generation GLM basic model and the "Fully Self-Training" system, as well as the deployment and upgrading of large-scale training, production inference, computing resources and related technical infrastructure;
- Approximately 15% is used for business expansion, strategic investments, and potential mergers and acquisitions, including minority equity investments or controlling acquisitions of companies or assets that are complementary to AI technology.
- Approximately 25% is used to optimize capital structure, supplement working capital, and for other general corporate purposes.

The company expects that the above funds will be fully utilized by June 30, 2028.
The announcement states that as the company continues to advance its research and development on next-generation GLM foundational models, fully self-trained systems, and long-term task reinforcement learning, and with the continued growth in demand for MaaS platforms, the demand for training and inference computing resources is correspondingly increasing. The company believes that "the supply and delivery conditions of high-quality computing resources in the market are currently quite favorable, but the process from signing a contract to deployment requires a certain period. Completing the financing arrangements at this stage will help align the deployment pace of new computing power with the company's expansion plans."
Impact of equity structure
Following this financing, the shareholding of the single largest shareholder group (composed of Beijing Lianpai Technology Development Center and other parties acting in concert) will be gradually diluted from the current approximately 28.58% (unlisted shares + H shares). Assuming the placement is completed and all bonds are converted, the group's total shareholding will decrease to approximately 25.89%, but the company's public float will still be higher than 10% of the enlarged total issued shares, which complies with the listing rules of the Stock Exchange of Hong Kong.
The company also undertakes that, from the date of the placement agreement to 60 days after the closing date, it shall not conduct any additional equity financing arrangements without the written consent of the joint placement agents.
Accelerated Fundraising: Two Moves Within Two Months
This is already one of Zhipu's intensive moves in the capital market.
According to Bloomberg, in July of this year, Zhipu raised approximately US$4 billion by issuing about 19.8 million shares at a price of HK$1,588 to HK$1,698 per share. At that time, computing power, research and development, and talent were also the main uses.
Going back further, in January of this year, Zhipu completed its IPO on the Hong Kong Stock Exchange, issuing approximately 37.42 million new H shares at HK$116.20 per share. The total net proceeds from the global offering (including the exercise of the over-allotment option) amounted to approximately HK$4.896 billion. The announcement shows that as of August 31, 2026, all IPO proceeds have been used, and approximately 34.92% of the proceeds from the July placement have been utilized.
With three rounds of financing, Zhipu has raised more than 10 billion US dollars from the capital market.

Industry Background: Competition in large-scale models has entered a capital-intensive phase.
From a broader perspective, Zhipu's intensive financing reflects the evolution of the competitive logic in the large-scale model industry.
The first round of competition in the large-scale model industry eliminated companies unable to continuously iterate their models. Today, a few players such as Zhipu, DeepSeek, Kimi, and MiniMax have proven their ability to remain at the model table. However, the things that need to be proven in the next round of competition are changing—not only the model capabilities themselves, but also the ability to obtain sufficient computing power, stably run large-scale clusters, integrate domestically produced chips, reduce training and inference costs, and most importantly, whether there is sufficient funding to continuously invest in all of this.
Models, talent, capital, and infrastructure are all indispensable. Zhipu's recent $5 billion funding round is largely a strategic move to prepare for the next phase of competition.
Risk Warning and DisclaimerInvesting involves risk; please exercise caution. This article does not constitute personal investment advice and does not take into account the specific investment objectives, financial situation, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article are suitable for their specific circumstances. Any investment decisions made based on this information are at your own risk.