Home / Technology / Zhipu AI's GLM-5 Challenges Global Tech Giants
Zhipu AI's GLM-5 Challenges Global Tech Giants
11 Feb
Summary
- Zhipu's GLM-5, with double parameters, targets coding and agentic tasks.
- The model is benchmarked directly against Anthropic's Claude Opus series.
- Zhipu's IPO this year has seen its stock jump over 50%.
Zhipu, a prominent AI developer based in Beijing, is preparing to launch its latest large language model, GLM-5, early on Thursday. This new iteration boasts more than double the parameters of its predecessor and is engineered to excel at intricate coding and agentic tasks. Its performance has been directly evaluated against Anthropic's Claude Opus series, highlighting a significant advancement in the company's AI capabilities.
The release of GLM-5 occurs amidst an increasingly competitive AI landscape, with industry peers like Moonshot AI and ByteDance also rolling out new models. Zhipu's strategic move ahead of the Lunar New Year anticipates further advancements, particularly with expectations that DeepSeek will soon unveil its next-generation architecture. This intense development cycle underscores the rapid pace of innovation within China's AI sector.
Since its inception in 2019, Zhipu, backed by Tsinghua University researchers and significant investment, has grown substantially, reaching an approximate valuation of $18 billion. The company is actively transitioning from providing customized AI solutions for domestic clients to offering its services globally. This expansion includes subscription plans for its GLM Coding Plan, aiming to reach international users.
