Zhipu AI launches open-source GLM-5.1 model for coding tasks

Z.ai has released GLM-5.1, an open-source model targeted at developer workflows and coding agents.

GLM-5.1 is designed to run autonomously on a single task for up to eight hours, handling planning, execution, and repeated optimization.

The company published benchmark results showing a 58.4 score on SWE-Bench Pro and released downloadable model weights in its open repo.

The model offers a 200K context window, 128K maximum output, streamed tool-call output, and availability to GLM Coding Plan users across Max, Pro, and Lite.