Meridiano

Add Article

No Image Available

Article Details

Title

Article: China’s Zhipu AI launches new major model GLM-5 in challenge to its rivals

Impact Score

5 / 10

AI Summary (Processed Content)

Chinese AI company Zhipu AI has launched its new flagship model, GLM-5, highlighting a significant leap in coding capabilities and a shift toward "agentic engineering." The announcement caused the company's Hong Kong shares to surge by 28.7 percent.

The company claims the model achieves leading scores for open models in coding and agentic performance, even surpassing Google's Gemini 3 Pro in internal tests, though it still lags behind Anthropic's Claude on coding benchmarks.

Technically, GLM-5 is much larger than its predecessor, with 744 billion parameters and trained on 28.5 trillion tokens, utilizing a new efficient architecture called DeepSeek Sparse Attention.

Main topics: Zhipu AI's GLM-5 model launch; its claimed performance and benchmarks; its technical specifications and architecture; the market reaction.

Original URL

https://www.scmp.com/tech/article/3343239/chinas-zhipu-ai-launches-new-major-model-glm-5-challenge-its-rivals?utm_source=rss_feed

Source Feed

Big Tech - South China Morning Post

Published Date

2026-02-11 21:22

Fetched Date

2026-03-04 14:30

Processed Date

2026-03-04 15:20

Embedding Status

Present

Cluster ID

Not Clustered

Raw Extracted Content

Advertisement
China’s Zhipu AI launches new major model GLM-5 in challenge to its rivals
The GLM-5 represents a shift in AI development from ‘vibe coding’ to ‘agentic engineering’ to generate an enhanced performance
3-MIN READ3-MIN
Listen
Chinese artificial intelligence company Zhipu AI has launched its latest flagship model, the GLM-5, with a leap in coding capabilities amid a heated race among Chinese tech firms to roll out major new models ahead of the Spring Festival holiday.
Zhipu AI, known internationally as Z.ai, said that the GLM-5 represented a shift in AI development from “vibe coding” to “agentic engineering” – AI-automated coding at a larger scale – thanks to its latest model’s enhanced performance.
The model’s release overnight pushed the Hong Kong shares of Zhipu, officially listed as Knowledge Atlas Technology, up 28.7 per cent to HK$402 on Thursday.
Advertisement
The company said that GLM-5 achieved industry-leading scores for open models in key areas such as coding and agentic performance according to internal tests, even surpassing Google DeepMind’s most powerful model, Gemini 3 Pro.
However, the model still lagged Anthropic’s Claude, a proprietary model and widely considered the world’s best coding model, across the board on coding benchmarks, according to the company’s self-reported scores.
Advertisement
The new model doubled the size of its predecessor, GLM-4.7, from 355 billion to 744 billion parameters and increased the training data to 28.5 trillion tokens. It also adopted a new architecture, first pioneered by the Hangzhou start-up DeepSeek, called DeepSeek Sparse Attention, designed to maximise computational and cost efficiency.
Advertisement
Select Voice
Select Speed
1.00x