智谱 GLM

GLM-5.1 is designed for long-horizon tasks, capable of working independently for up to 8 hours in a single run, covering planning, execution, refinement, and delivery. It demonstrates strong engineering intelligence in autonomous planning, sustained execution, bug fixing, and strategy iteration, aligning with Claude Opus 4.6 capabilities. Built with multi-turn SFT, RL, and a process-quality evaluation framework, it improves stability, consistency, and tool use over extended tasks.