# China%20AI
Latest news and articles about China%20AI
Total: 30 articles found
China’s StepFun Taps Qianli Chief and Raises >RMB5bn to Push Foundation Models into Devices
StepFun has appointed Qianli Technology chairman Yin Qi as its own chairman and closed a B+ round exceeding RMB 5 billion to accelerate development of foundation models and commercialise AI on devices. The funding and leadership change underscore a strategic push in China to embed large models into cars, smartphones, wearables and robots through close hardware partnerships.
Old Guard Returns: Yin Qi’s Dual Chairmanship and a RMB5bn Vote of Confidence Reshape China’s AGI Race
Yin Qi has been appointed chairman of StepFun while retaining the chairmanship of Qianli Technology, as StepFun closes a B+ round exceeding RMB5 billion. The move pairs deep foundation-model R&D with a hardware-centred commercialization strategy — notably automotive — and signals a new phase of consolidation and specialization in China’s AI landscape.
Zhipu Limits Sales of GLM Coding Plan to Protect Long‑standing Users After GLM‑4.7 Demand Spike
Zhipu has limited daily sales of its GLM Coding Plan to 20% of previous volumes as a temporary measure after GLM‑4.7 triggered heavy demand that strained compute resources and slowed model responses during peak hours. The cap, beginning Jan 23 and refreshed daily, aims to protect existing users while Zhipu expands capacity and tightens control over malicious traffic.
Zhipu AI Temporarily Caps GLM Coding Plan Sales After GLM‑4.7 Triggers Surge in Demand
Zhipu AI will temporarily cap daily sales of its GLM Coding Plan to 20% of current levels starting January 23 after the GLM‑4.7 release caused spikes in usage and intermittent slowdowns during weekday peak hours. The move protects existing subscribers and highlights persistent infrastructure and cost challenges for providers of code‑focused large language models.
Fu Sheng: AI Will Make Machines Orbit People — Robots, Agents and the Case for ‘Start with the End’
Fu Sheng, founder of Cheetah Mobile and OrionStar, argues the AI era will flip human–machine roles and calls for rebuilding apps as AI‑native products. He stresses a scenario‑first approach for robots and devices to secure commercial returns, while warning that foundational models still need improvement for decision‑critical tasks.
Small, Open and Multimodal: Chinese Startup Releases 10‑Billion‑Parameter Vision‑Language Model Claiming SOTA Performance
Chinese start‑up Jieyue Xingchen open‑sourced Step3‑VL‑10B, a 10‑billion‑parameter multimodal model that the team says matches same‑scale state‑of‑the‑art performance on vision, reasoning, math and dialogue. The release highlights a push for efficient, deployable multimodal models and will prompt independent verification and community adoption.