Claude Opus 4.6
We’re upgrading our smartest model.
A named Claude Opus model release from Anthropic for language-model tasks, including use in Claude Code.
Pricing
Official site · Nov 8, 2026, 7:42 AM
Input / 1M
$5.00
Output / 1M
$25.00
Cached input / 1M
$0.50
Standard global API rates; cache writes: 5-minute $6.25 and 1-hour $10.00 per 1M tokens; cache hits/refreshes $0.50 per 1M. US-only standard inference costs $5.50 input / $27.50 output. Global batch processing costs $2.50 input / $12.50 output. Prices effective 2026-09-28.
Anthropic's official list-price card effective September 28, 2026 lists Claude Opus 4.6 Claude API standard global pricing at $5.00 per 1M input tokens and $25.00 per 1M output tokens. US-only inference is $5.50 input and $27.50 output; batch processing global is $2.50 input and $12.50 output. Reported base rates are standard global.
Model Intelligence
Arena ranking
26
Benchmarkable
Yes
Model level
release
GPQA
0.84
HLE
0.19
IFBench
0.45
LCR
0.67
TerminalBench Hard
0.48
TAU2
0.85
Recent stories
1 linked story