Zhipu AI Secures Strategic Financing from Hangzhou State-Owned Investors
On March 3rd, Zhipu AI, one of China’s leading AI companies known as the “Six Tigers”, announced a strategic financing round exceeding 1 billion RMB, with participation from state-owned investors in Hangzhou.
This funding aims to drive innovation and ecosystem development for Zhipu AI’s domestic GLM large model. The company plans to make 2025 its open-source year, with a new large model set for release soon.
In December 2024, Zhipu AI completed a financing round of around 3 billion RMB, revealing impressive achievements such as over 100% revenue growth in commercialization and a 150-fold increase in daily platform token consumption.
Zhipu AI has secured 13 rounds of financing to date, with major investors from Beijing, including companies from Zhongguancun Science City and the Beijing Artificial Intelligence Industry Investment Fund. A significant financing round in September 2024 helped Zhipu AI reach a valuation of 20 billion RMB.
The rise of DeepSeek has caused significant shifts in the domestic AI market among the “Six Tigers,” prompting a focus on commercialization and open-source strategies.
Zhipu AI has traditionally operated with a dual B2B and B2C business model. In the B2B sector, Zhipu offers API call charges, cloud privatization, and on-premises privatization. The company has also partnered with major players like Samsung and Honor, embedding its large model in the system layer of Samsung’s Galaxy S25 series for the Chinese market.
On the B2C side, Zhipu relies on subscription fees from its Zhipu Qingyan app, which boasted 7.02 million monthly active users in January 2025.
DeepSeek’s success has influenced strategic shifts in other major large model companies. For instance, Kimi, known for its heavy marketing investment strategy, is reportedly reassessing this approach to focus more on basic model training.
Moonshot is emphasizing State-of-the-Art (SOTA) basic models and may retrain these models to capitalize on application layer opportunities, including the development of DeepResearch for deep research fields.
01.AI is likely seeking growth in the B2B sector, with CEO Kai-fu Lee indicating a shift in strategy following the release of the Yi-Large model last year.
MiniMax recently open-sourced the new 01 series model, supporting up to 4 million tokens for context input. Bai Chuan Intelligence is concentrating on applying large models in the medical field, with plans to launch five AI medical products in collaboration with the National Children’s Medical Center, Beijing Children’s Hospital.
As of now, 01.AI, MiniMax, and Jieyue Xingchen have integrated with DeepSeek, while Zhipu AI has announced that 2025 will be their year of open source.
