An

claude-opus-5-5
Benchmark & Insights

Anthropic Claude API
Updated Oct 4, 2026 All models
Sample size
232 runs
in window
Accuracy
75.5%
reference match · 282d
Confidence
83%
over 282 runs
Window end
Oct 4, 2026
most recent run
Input price
$4.00/MTok
prompt tokens
Output price
$20.00/MTok
completion tokens
Model insights
  • 01 Optimistic: it rated risk too low on 41 days and called 35 "MEDIUM_SAFE" days "low".
  • 02 It missed 13 of 25 "unsafe" days, including both late-September "HIGH" days, though it never raised a false alarm.
  • 03 A small step up from claude-opus-5 at a lower $4/$20, but claude-sonnet-5-5 beats it at half the price.
Notes

Introduced on September 22, 2026 20% cheaper than Claude Opus 5. Has a breaking changes comparing to a previous version.

Recent forecasts