The Bleeding Edge
← Back to the timeline

June 19, 2026

Zhipu AI (GLM)model

GLM 5.2 becomes the best open-weight coding model — MIT-licensed, self-hostable

Zhipu/Z.ai's GLM 5.2 (753B MoE, 40B active, 1M-token context) scores 62.1 on SWE-bench Pro, just behind Claude Opus 4.8, with open weights and GGUF/llama.cpp local support — frontier-class coding with no API bill and no US lab.