log inget an API key
Platform benchmark

claude-opus-5

anthropic · GPQA-Diamond · seed 42 · 2026-11-02 05:32 UTC

ResultsGPQA-Diamond · n=96 · seed 42
92.7%

92.7% correct · 99.0% answered on GPQA-Diamond

Score per $
0.36
Run cost
$2.61
Latency p50
11.8 s
Latency p95
65.9 s
Items
96/96 completed
Dropped
0

Dataset: GPQA-Diamond · question and response text withheld under dataset access terms.