Zhipu’s ARR growth outpaces Anthropic; China model downloads account for 41%, surpassing the U.S. for the first time

ZHIPU AI-25.32%
TENCENT-5.04%
GLM-2.90%

According to 36Kr on July 17, based on multiple independent sources, as of July, Zhipu (02513.HK)’s annual recurring revenue (ARR) has reached $1 billion. Zhipu increased from $100 million to $1 billion in 5 months, faster than Anthropic. Citing the platform’s latest data, Hugging Face said that China’s open-source models account for 41% of monthly and total download shares, surpassing the United States for the first time.

Zhipu’s ARR hits $1 billion in five months from $100 million

Based on data shared by 36Kr from multiple independent sources, as of July 2026, Zhipu ARR reached $1 billion. It took only 5 months to grow from $100 million to $1 billion, faster than Anthropic’s 15 months. 36Kr described Zhipu as China’s first foundation model company to cross the $1 billion ARR threshold; as of the report date, Zhipu had not responded to an inquiry for comment.

After its core product, GLM Coding Plan, cumulatively increased prices by 83%, demand still outstrips supply. By the end of the first quarter of 2026, the number of paid developers had reached 240,000. Zhipu also released an open-source model, GLM-5.2, built specifically for agentic coding tasks. According to the original report, in multiple benchmark tests, its performance is already on the same competitive level as Anthropic’s latest closed-source model.

For industry comparison, Anthropic’s Claude Code surged to $1 billion in ARR half a year after its release, driving Anthropic’s overall valuation to reach $1 trillion in May 2026.

China’s open-source model download share hits 41%, surpassing the U.S. for the first time

Citing the platform’s latest data, Hugging Face CEO Clem Delangue said that China’s open-source models account for 41% of monthly and overall download shares, surpassing the United States for the first time and emerging as the main force behind the global open-source ecosystem. Hugging Face currently hosts nearly 3 million public models and 1 million public datasets; more than half of the world’s Fortune 500 companies have deployed private or open-source models on the platform.

On the leaderboard of the AI model calling platform OpenRouter, the current top six most popular models are all developed by Chinese institutions:

· Tencent

· Xiaomi

· DeepSeek

· MiniMax

· Zhipu (two models in the list)

Anthropic Claude Opus 4.7 is currently ranked seventh on OpenRouter.

Vercel and enterprise adoption data

According to statistics from the Vercel platform, as of June 2026, open-source models processed nearly one-third of AI requests on the platform, taking on a large share of infrastructure-intensive workloads.

In a public analysis, Hugging Face CEO Delangue pointed out that tech companies and developers are realizing that core capabilities should not be outsourced to uncontrollable “black-box APIs.” Open-source models allow enterprises to host and fine-tune on their own infrastructure, reducing maintenance costs and ensuring data sovereignty.

FAQ

Where does the Zhipu ARR data come from, and has the company responded?

According to a report published by 36Kr on July 17, 2026, based on multiple independent sources, as of July 2026, Zhipu’s ARR had reached $1 billion and grew from $100 million to $1 billion in 5 months; as of the report date, Zhipu had not responded to the inquiry for comment.

What is the download share of China’s open-source models on Hugging Face?

According to data cited by Hugging Face CEO Clem Delangue from the platform, China’s open-source models account for 41% of monthly and overall download shares, surpassing the United States for the first time and becoming the main force in the global open-source ecosystem.

Which institutions are behind the current top six popular models on OpenRouter?

According to OpenRouter ranking data, the top six popular models are all developed by Chinese institutions, including Tencent, Xiaomi, DeepSeek, MiniMax, and Zhipu (two models in the list); Anthropic Claude Opus 4.7 is currently ranked seventh.

Disclaimer: The information on this page may come from third-party sources and is for reference only. It does not represent the views or opinions of Gate and does not constitute any financial, investment, or legal advice. Virtual asset trading involves high risk. Please do not rely solely on the information on this page when making decisions. For details, see the Disclaimer.
Comment
0/400
No comments