Zhipu, one of China’s most prominent AI startups, posted a 132% revenue jump in 2025, according to The Information. It’s a striking number that puts the Beijing-based company firmly among the fastest-growing players in the global AI race.
The details beyond the topline figure remain thin, but the signal is clear: Chinese AI companies aren’t just building impressive models. They’re turning them into real businesses.
Why this matters
Zhipu (also known as Zhipu AI or 智谱) was founded in 2019 as a spinoff from Tsinghua University’s research labs. The company builds large language models, with its GLM series competing against both domestic rivals like Baidu’s Ernie and Moonshot AI, and global heavyweights like OpenAI and Anthropic.
A 132% revenue increase suggests Zhipu is finding serious traction with enterprise customers in China. That’s notable for a few reasons:
- Enterprise AI adoption in China is accelerating. While U.S. companies dominate the headlines, Chinese firms are rapidly deploying AI across manufacturing, finance, and government services.
- Monetization is the hard part. Many AI startups globally still struggle to convert model capabilities into sustainable revenue. Zhipu crossing this threshold at scale is significant.
- The competitive landscape is brutal. China’s AI market has dozens of well-funded startups, plus tech giants like Alibaba, Baidu, and Tencent all pushing their own models. Growing revenue this fast means Zhipu is winning deals against serious competition.
Context: China’s AI funding boom
Zhipu has raised over $4 billion in funding, with backers including Saudi Arabia’s Prosperity7 and major Chinese tech firms. That puts it among the best-capitalized AI startups anywhere in the world.
The revenue growth also comes at a time when Chinese AI companies face a unique set of pressures. U.S. export controls on advanced chips have forced Chinese firms to get creative with training infrastructure, often relying on older Nvidia hardware or domestic alternatives from Huawei. Despite these constraints, companies like Zhipu and DeepSeek have delivered competitive models.
What to watch
The big question now is profitability. Revenue growth of 132% is impressive, but AI infrastructure is expensive. Training and serving large models requires massive compute budgets. Whether Zhipu can turn this revenue growth into a sustainable business, or whether it’s burning through its war chest to buy market share, will determine its long-term trajectory.
For the broader AI industry, Zhipu’s numbers reinforce a trend: the AI market is becoming genuinely global. The assumption that U.S. companies would dominate AI commercialization is looking increasingly outdated.
More details are available at The Information.