Realtime AI News
Zhipu AI nears 7 million API users, activates over 50,000 domestic compute chips
Zhipu AI's API user base has approached 7 million, and the company has newly activated more than 50,000 domestic compute chips, according to Wall Street CN. The dual milestone highlights growth in both its developer ecosystem and its push for compute self-reliance.
Zhipu AI's API user count has approached 7 million, and the company has newly activated more than 50,000 domestically produced compute chips, according to a Wall Street CN report.
The two figures mark fresh progress on both the application and infrastructure fronts: a growing API ecosystem means more developers and enterprises are plugging into Zhipu's models, while the large-scale chip deployment points to a more self-reliant compute supply chain.
Zhipu is one of China's leading large-model developers, serving models to developers and enterprises through its API; the rapid growth of its user base reflects how quickly domestic models are being adopted in real applications.
Activating more than 50,000 domestic chips signals heavier reliance on homegrown hardware for training and inference, which is also a positive development for China's domestic chip supply chain.
For the industry, simultaneous growth in API users and compute capacity suggests leading Chinese model vendors are shifting the contest from raw model capability toward ecosystems and infrastructure.
The report does not disclose which chip suppliers are involved, where the chips are deployed, or how capacity is split between training and inference — details that still await official clarification.
Watch for Zhipu to publish more detailed compute plans, and for signs of paid conversion among enterprise and developer API customers.
Why it matters
Zhipu's near-7-million API users and large-scale activation of domestic chips show Chinese large-model vendors accelerating on both ecosystem and compute self-reliance, shifting competition toward infrastructure.
Nearby Updates
All08/10, 22:54
KYND taps Anthropic AI to test cyber risk
KYND, a cyber risk management company, has turned to Anthropic's AI to test cyber risk, according to FinTech Global. The move adds another enterprise deployment of Anthropic models in the security assessment space.
08/10, 23:05
Weathernews Launches Multilingual AI Agent to Help Businesses Manage Weather Risks
Weathernews has added a chat-based, multilingual AI agent to its corporate service Weathernews for Business to support operational decision-making under weather risk. Built on four decades of weather-risk know-how, the agent handles typhoon response planning, heatstroke-risk guidance, and queries such as when WBGT will reach severe caution levels.
08/10, 22:30
Google brings new AI and agentic tools to Ads and Analytics
Google announced a new wave of AI and agentic features across Google Ads and Google Analytics, aimed at simplifying marketing workflows through automation. The company's official blog says the updates help advertisers and analysts handle campaign optimization and data insights more efficiently.
08/10, 22:00
OpenAI Writes to Texas Governor Abbott on Responsible AI Infrastructure
OpenAI has sent a letter to Texas Governor Greg Abbott outlining its commitment to responsible AI infrastructure in the state, saying the company supports reliable, transparent growth that benefits Texans. The letter marks a direct outreach to state leadership as AI infrastructure siting and permitting become increasingly contested policy issues.