Op
gpt-6.1-sol
Benchmark & Insights
OpenAI OpenAI API
Updated Sep 30, 2026 All models
Sample size
229 runs
in window
Accuracy
88.2%
consensus match · 229d
Confidence
87%
over 229 runs
Window end
Sep 30, 2026
most recent run
Input price
$2.00/MTok
prompt tokens
Output price
$10.00/MTok
completion tokens
Model insights
- 01 Always errs on the cautious side: it never rates risk too low, and nearly every miss is a "medium" call on a "low" day.
- 02 About two-thirds of its misses fall on the same days as gpt-6-sol's, with slightly fewer jumps to "high", so the 6.1 update changed little.
- 03 It never called an "unsafe" day "safe".
Recent forecasts