Op
gpt-5-mini Benchmark & Insights
OpenAI OpenAI API
Updated Jun 6, 2026 All models
Sample size
113 runs
in window
Accuracy
87.0%
consensus match · 154d
Confidence
80%
over 158 runs
Window end
Jun 6, 2026
most recent run
Input price
$0.25/MTok
prompt tokens
Output price
$2.00/MTok
completion tokens
Model insights
- 01 Two failure modes in one model: 6 "safe" calls on consensus "unsafe" days early in the record, then a long run of "medium" overcalls on "low" days from March onward.
- 02 Very cheap ($0.25/$2.00), and it still outscores the 3x-pricier gpt-5.4-mini.