Claude 3.7 Sonnet

anthropic · Ranked across 5 benchmarks · best rank #3

Benchmark scores

BenchmarkCategoryRankScoreCaptured
METR Task Horizon (HCAST) agents #3 1h3m 2025-07-12
Aider Polyglot code #9 64.9% 2025-02-24
SWE-bench Verified agents #17 66.4% 2025-05-14
OpenRouter · Weekly Usage usage #32 #103 2026-05-02
Chatbot Arena chat #41 1314 2026-04-30