GLM-5.1: Towards Long-Horizon Tasks
Open paper
Benchmark
Model
Rank
Results
code-generation-on-swe-bench-pro
GLM-5.1
#6
Resolved: 58.4
code-generation-on-terminal-bench-2-1
Terminus-2 + GLM-5.1
#8
Accuracy: 63.5
Rank counts only results with a code link.