Platform benchmark

claude-opus-4-8

anthropic · MATH-500 · seed 42 · 2026-07-16 00:03 UTC

ResultsMATH-500 · n=96 · seed 42
92.7%

92.7% correct · 100.0% answered on MATH-500

Score per $
1.14
Run cost
$0.81
Latency p50
3.0 s
Latency p95
10.5 s
Items
96/96 completed
Dropped
0