Claude Opus 4

anthropic · Ranked across 6 benchmarks · best rank #2

Benchmark scores

BenchmarkCategoryRankScoreCaptured
AA Coding Index code #2 74.3% 2026-07-31
OpenRouter · Weekly Usage usage #3 #9 2026-06-09
Aider Polyglot code #7 72.0% 2025-05-25
Chatbot Arena chat #9 1463 2026-07-27
SWE-bench Verified agents #10 73.2% 2025-05-22
METR Time Horizon agents #12 1h40m 2025-05-22