ARTFEED — Contemporary Art Intelligence

Zhipu AI's GLM-5.2 ranks top three globally on major benchmark

ai-technology · 2026-06-26

On June 13, Zhipu AI (Z.ai), a Chinese artificial intelligence startup, unveiled GLM-5.2, an economical coding model. This model has achieved a remarkable milestone by becoming the first Chinese AI to secure a spot among the top three globally on a significant benchmark, outpacing recent models such as DeepSeek V4 Pro, MiniMax M3, and Alibaba's Qwen3.7-Max. Matt Velloso, a former VP at Meta and Google DeepMind, commended GLM-5.2 as a reliable daily tool, highlighting its brevity compared to OpenAI's GPT-5.5. This announcement followed Anthropic's decision to pause Claude Fable 5 to adhere to US regulations regarding foreign users.

Key facts

  • Zhipu AI released GLM-5.2 on June 13.
  • GLM-5.2 is the first Chinese model to rank top three globally on a major benchmark.
  • Matt Velloso, former VP at Meta and Google DeepMind, praised GLM-5.2 as a daily driver.
  • Velloso compared GLM-5.2 favorably to OpenAI's GPT-5.5.
  • Anthropic shelved Claude Fable 5 a day before GLM-5.2's launch due to US restrictions.
  • Other Chinese models include DeepSeek V4 Pro, MiniMax M3, and Alibaba's Qwen3.7-Max.
  • GLM-5.2 is known internationally as Z.ai.

Entities

Institutions

  • Zhipu AI
  • Z.ai
  • Meta Platforms
  • Google DeepMind
  • OpenAI
  • Anthropic
  • Alibaba

Locations

  • China
  • United States

Sources