Op
gpt-5.5 Benchmark & Insights
OpenAI OpenAI API
Updated Jun 6, 2026 All models
Sample size
43 runs
in window
Accuracy
97.8%
consensus match · 45d
Confidence
88%
over 45 runs
Window end
Jun 6, 2026
most recent run
Input price
$5.00/MTok
prompt tokens
Output price
$30.00/MTok
completion tokens
Model insights
- 01 The best model with a real sample: one lone divergence in 43 days, a cautious "medium" on a consensus "low" day.
- 02 It fixes both the optimism of gpt-5.1 and the overcaution of gpt-5.4, but it is also the most expensive model on the panel ($5/$30 per MTok).
Notes
Breakpoint pricing at 272K tokens (doubles to $10/$45 above threshold)