Navigate
AI BENCHY
Advertise here

DeepSeek V3.2 vs Hy4 preview

DeepSeek V3.2 leads on average score with 5.0 vs 4.8. Hy4 preview has the lower benchmark cost at $0.041 vs $0.054. DeepSeek V3.2 is faster at 17.89s vs 39.24s, with pass rates of 36.4% vs 16.7%.

Last updated at: 2026-08-29

Rank
#248
Total Output Tokens
42,099
Response Time (avg)
17.89s
Total Cost
$0.054
Rank
#253
Total Output Tokens
7,306
Response Time (avg)
39.24s
Total Cost
$0.041
Recommended model DeepSeek V3.2

It has the best score here (5.0), while responding about 2.2x faster than Hy4 preview.

Detailed comparison

Metric DeepSeek V3.2 DeepSeek V3.2 none Release: 2025-12-01 Hy4 preview Hy4 preview none Release: 2026-08-29
Score 5.0 4.8
Rank #248 #253
Reliability 10.0 4.3
Consistency 8.1 9.3
Attempts 66/66 66/66
Tests Correct
Attempt pass rate 36.4% 16.7%
Flaky tests 5 1
Total Runs 66 66
Cost per result 0.870 1.343
Total Cost $0.054 $0.041
Input Price $0.269 / 1M $0.834 / 1M
Output Price $0.400 / 1M $2.501 / 1M
Total Input Tokens 135,831 26,376
Output Tokens 42,099 7,306
Reasoning Tokens 0 0
Response Time (avg) 17.89s 39.24s
Response Time (max) 115.89s 95.32s
Response Time (total) 393.53s 863.30s
Parameters 671B total (37B active) 770B total (49B active)
Availability Open source Open source

Model generation showcase

Hamster playing table tennis

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

#248 DeepSeek V3.2

none
Cost
$0.002
Time
7.0s
Tokens
1,046 tok

#253 Hy4 preview

none
This operation was aborted
Cost
$0.000
Time
0.0s
Tokens
0 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
DeepSeek V3.2 3.1 6.9 11.1% 1 14.54s 7,279 4,528 0
Hy4 preview 3.8 9.4 0.0% 0 45.53s 6,646 4,103 0

Quick Compare

Switch Comparison Pair