An
claude-haiku-4-5 Benchmark & Insights
Anthropic Claude API
Updated Jul 31, 2026 All models
Sample size
169 runs
in window
Accuracy
92.7%
consensus match · 206d
Confidence
92%
over 207 runs
Window end
Jul 31, 2026
most recent run
Input price
$1.00/MTok
prompt tokens
Output price
$5.00/MTok
completion tokens
Model insights
- 01 Matches Claude Opus 5 to the decimal at a fifth of the price, which makes it the sensible Claude default.
- 02 It leans optimistic with repeat "low" calls on "medium" days, yet it still reached for "high" more often than any Claude except the badly calibrated Sonnet 4.6.
Recent forecasts