log inget started
Platform benchmark

gpt-5.4-nano

openai · GPQA-Diamond · seed 42 · 2026-07-15 13:52 UTC

ResultsGPQA-Diamond · n=96 · seed 42
54.2%

54.2% correct · 100.0% answered on GPQA-Diamond

Score per $
6.69
Run cost
$0.08
Latency p50
5.3 s
Latency p95
10.7 s
Items
96/96 completed
Dropped
0