Navigate
AI BENCHY
Advertise here

Anthropic: Claude Sonnet 4.6 vs DeepSeek: DeepSeek V3.2

Claude Sonnet 4.6 leads on average score with 7.3 vs 7.0. DeepSeek V3.2 (medium) has the lower benchmark cost at $0.078 vs $0.661. Claude Sonnet 4.6 is faster at 8.12s vs 68.62s, with pass rates of 57.6% vs 65.2%.

Recommended modelClaude Sonnet 4.6It has the best score here (7.3), while responding about 8.5x faster than DeepSeek V3.2 (medium).

Last updated at: 2026-07-25

Metric Claude Sonnet 4.6 Claude Sonnet 4.6 none Release: 2026-02-17 DeepSeek V3.2 DeepSeek V3.2 medium Release: 2025-12-01
Score 7.3 7.0
Rank #71 #84
Reliability 10.0 10.0
Consistency 9.7 7.4
Tests Correct
Attempt pass rate 57.6% 65.2%
Flaky tests 1 7
Total Runs 66 66
Cost per result 5.502 0.671
Total Cost $0.661 $0.078
Input Price $3.000 / 1M $0.269 / 1M
Output Price $15.000 / 1M $0.400 / 1M
Total Input Tokens 123,264 101,047
Output Tokens 19,362 11,834
Reasoning Tokens 0 117,014
Response Time (avg) 8.12s 68.62s
Response Time (max) 51.18s 376.10s
Response Time (total) 121.78s 1509.53s

Hamster playing table tennis

Prompt: Create a detailed SVG illustration of a hamster playing table tennis.

#71 Claude Sonnet 4.6

none
Cost
$0.038
Time
27.3s
Tokens
2,598 tok

#84 DeepSeek V3.2

medium
Cost
$0.001
Time
53.6s
Tokens
1,932 tok

Top Models by Score

Score vs Total Cost

Response Time (avg)

Score vs Response Time (avg)

Total Output Tokens

Score vs Total Output Tokens

Category Breakdown

Coding Score Consistency Attempt pass rate Flaky tests Tests Correct Response Time (avg) Input Tokens Output Tokens Reasoning Tokens
Claude Sonnet 4.6 5.5 10.0 33.3% 0 5.19s 8,522 2,127 0
DeepSeek V3.2 6.0 7.2 55.6% 1 248.68s 5,717 649 52,014

Quick Compare

Switch Comparison Pair