Zhipu AI open-sources GLM-5.2 model with 1M context, tops Code Arena for available models globally
1 hour ago
2 languages2 countries2 sources
AI summary
On June 17, Zhipu AI released and open-sourced the GLM-5.2 model. According to VentureBeat, the 753-billion parameter model is claimed to be the top-ranked globally available model on the Code Arena benchmark, integrating 1 million tokens of lossless context. The company states the model is optimized for long-context tasks and coding performance.
Why this matters
- Major story by our editors' read
Key facts
- GLM-5.2 is an open-source model released by Z.ai (formerly Zhipu AI).Confirmed by multiple sources · IT之家, VentureBeat
- The model has 753 billion parameters and supports a 1M token context window.Confirmed by multiple sources · IT之家, VentureBeat
- On the Code Arena benchmark, GLM-5.2 achieved the highest rank among globally available models.Reported by a single source · IT之家
- GLM-5.2 outperforms GPT-5.5 on multiple long-horizon coding benchmarks at a fraction of the cost.Reported by a single source · VentureBeat
- The model is available on Hugging Face, the Z.ai API, and over 20 third-party coding environments.Reported by a single source · VentureBeat
- GLM-5.2 is optimized to run on domestic Chinese computing platforms from day zero.Reported by a single source · IT之家

Timeline
IT之家· China
智谱开源 GLM-5.2 模型:1M 无损上下文,Code Arena 上全球可用模型第一
VentureBeat· United States
Z.ai’s open-weights GLM-5.2 beats GPT-5.5 on multiple long-horizon coding benchmarks for 1/6th the cost