invokeable()

Claude Opus 5 vs GLM-5.1

A custom comparison card. Numbers link back to their entries; data as of 2026-08-19. Build your own →

model
Claude Opus 5
Anthropic
model
GLM-5.1
Zhipuai
Context1M200K
Input $/M$0.98
Output $/M$3.08
Knowledge cutoff2026-05
Available viaAnthropic APIOpenRouter
SWE-bench Verified (resolved)96
SWE-bench Pro79.2%
SWE-bench Multilingual89.5%
SWE-bench Multimodal59.4%
DeepSWE68.8%
FrontierCode (mean@5)53.4
Artificial Analysis Coding Agent Index (average pass@1)52.7
SWE-Atlas Codebase QnA (pass@1)73.2
SWE-bench Pro (pass@1)19.8
Terminal-Bench (pass@1)65.1