An
claude-sonnet-4-6
Benchmark & Insights
Anthropic Claude API
Updated Oct 4, 2026 All models
Sample size
232 runs
in window
Accuracy
76.9%
reference match · 251d
Confidence
88%
over 253 runs
Window end
Oct 4, 2026
most recent run
Input price
$3.00/MTok
prompt tokens
Output price
$15.00/MTok
completion tokens
Model insights
- 01 The most cautious Claude: it leans high, rating 22 "LOW" days "medium", and flags 12 of 19 "MEDIUM_UNSAFE" days as "unsafe".
- 02 It was the best model over the Aug-Oct stretch, ahead of both Fable models.
- 03 Weak in March (57%), a month packed with hard "MEDIUM_UNSAFE" days.
Recent forecasts