Model WarsJune 17, 2026via The Decoder
Zhipu AI's GLM-5.2 closes in on closed-source leaders in coding marathons
Why it matters
Zhipu AI's GLM-5.2 demonstrates that open-source models are rapidly closing the capability gap with closed-source leaders on specialized benchmarks like coding marathons, while remaining available under permissive licensing—reshaping competitive dynamics in the model hierarchy.
Key signals
- GLM-5.2 achieves 1M stable context window
- Released under MIT license (open-source)
- Trails Claude Opus 4.8 by just 1 percentage point on FrontierSWE benchmark
- FrontierSWE measures hours-long coding tasks
- Still lags closed-source models on reasoning benchmarks
- Chinese AI lab (Zhipu AI)
The hook
Open-source just got 1 percentage point away from Claude Opus on coding tasks. Zhipu AI's GLM-5.2 is the narrowing gap leaders need to watch.
Chinese AI lab Zhipu AI releases GLM-5.2 with a stable 1-million-token context under the MIT license. On FrontierSWE, a benchmark for hours-long coding tasks, the open-source model trails Anthropic's Claude Opus 4.8 by just one percentage point. On reasoning, it still falls well behind closed-source rivals.