Zhipu AI (also known as Zhipu or BigModel) was founded in 2019 as a spin-off from the Knowledge Engineering Group at Tsinghua University, one of China’s premier research institutions. The company has become one of the leading Chinese AI labs, competing directly with DeepSeek and Alibaba Qwen for domestic enterprise market share.
Core Products
GLM-4 is Zhipu’s flagship language model, designed with a General Language Model architecture that differs from standard GPT-style autoregressive models. ChatGLM is the consumer-facing chatbot built on GLM. CogView and CogVideo provide image and video generation capabilities, making Zhipu one of the few Chinese labs with strong multimodal offerings across text, image, and video.
Competitive Position
Zhipu’s academic roots give it strong connections to the Chinese research community and access to talent from Tsinghua. The company has attracted investment from virtually every major Chinese tech company, suggesting broad industry support. However, Zhipu faces intense competition from DeepSeek’s efficiency innovations and Alibaba Qwen’s open-source dominance.
Recent Developments
Zhipu raised over $340 million in 2024, reaching a $3 billion valuation. GLM-4 showed significant improvements in reasoning and coding, though benchmarks remain somewhat behind DeepSeek-V3 and Qwen 2.5. The company has expanded its enterprise offerings and API platform while continuing multimodal research with CogView and CogVideo updates.
Outlook
Zhipu is well-positioned within the Chinese AI market due to its academic pedigree, diverse investor base, and multimodal capabilities. The company’s challenge is differentiating in an increasingly crowded Chinese AI market where DeepSeek has captured international mindshare and Alibaba commands enterprise distribution. Zhipu’s strength lies in comprehensive platform offerings rather than any single model breakthrough.