Claude Sonnet 4.6 vs DeepSeek V4 Pro

DeepSeek V4 Pro edges ahead on overall intelligence (RunFree 72.4 vs 66.2). Here's how they stack up on benchmarks, price and specs.

Claude Sonnet 4.6
Anthropic
DeepSeek V4 Pro
DeepSeek

Shared benchmarks (4)

Loading chart…
Claude Sonnet 4.6DeepSeek V4 Pro
RunFree Score66.272.4
Blended price / 1M$6.00$0.54
Input / 1M$3.00$0.43
Output / 1M$15.00$0.87
Context window1M1.0M
Max output128K384K
Reasoning modelYesYes
GPQA Diamond89.9%90.1%
Humanity's Last Exam33.2%37.7%
SWE-Bench Verified79.6%80.6%
Terminal-Bench59.1%67.9%
Wins010

A blank means the metric is tied or one model has no verified data. Prices are blended 3:1 input:output.

Compare more: Claude Sonnet 4.6 card · DeepSeek V4 Pro card · full leaderboard