Agents3h ago
GLM-5.2 Sets New Open Agent Benchmark
The GLM-5.2 model hits a critical capability threshold for open-source agentic workflows. It demonstrates a sharp jump in reliability for complex tool use and autonomous task execution. This release narrows the gap between proprietary systems and open weights. Developers can now deploy more stable agents without relying on closed APIs.