Z.AI Introduces GLM-5.1: An Open-Weight 754B Agentic Model That Achieves SOTA on SWE-Bench Pro and Sustains 8-Hour Autonomous Execution

Z.AI released GLM-5.1, its flagship open-weight 754B agentic coding model.

The model uses a Mixture-of-Experts plus DSA architecture and asynchronous RL training for stronger tool use and long-horizon autonomy.

Z.AI says GLM-5.1 outperforms GLM-5 and delivers state-of-the-art results on SWE-Bench Pro.

It also shows gains on agent, browsing, and terminal benchmarks including Terminal-Bench 2.0 and NL2Repo.

The launch underscores continued progress in open-weight AI for software engineering workflows.