Navigate
AI BENCHY
Your ad here

AI BENCHY Compare

DeepSeek: DeepSeek V3.2 vs DeepSeek: DeepSeek V4 Pro

Last updated at: 2026-04-24

Metric DeepSeek V3.2 DeepSeek V3.2 none Release: 2025-12-01 DeepSeek V4 Pro DeepSeek V4 Pro none Release: 2026-04-24
Score 6.0 6.7
Rank #72 #59
Consistency 8.6 9.5
Tests Correct
Attempt pass rate 46.3% 51.9%
Flaky tests 3 1
Total Runs 52 26
Cost per result 0.225 0.317
Total Cost $0.016 $0.029
Input Price $0.252 / 1M $1.740 / 1M
Output Price $0.378 / 1M $3.480 / 1M
Output Tokens 8,378 1,596
Reasoning Tokens 0 0
Response Time (avg) 12.07s 24.23s
Response Time (max) 115.89s 109.46s
Response Time (total) 217.28s 436.17s

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

Anti-AI Tricks Score Consistency Attempt pass rate Flaky tests Tests Correct Response Time (avg) Output Tokens Reasoning Tokens
DeepSeek V3.2 3.2 9.8 0.0% 0 7.63s 1,419 0
DeepSeek V4 Pro 4.8 10.0 25.0% 0 36.12s 221 0
Coding Score Consistency Attempt pass rate Flaky tests Tests Correct Response Time (avg) Output Tokens Reasoning Tokens
DeepSeek V3.2 2.4 1.3 33.3% 1 7.63s 553 0
DeepSeek V4 Pro 10.0 10.0 100.0% 0 33.40s 246 0
Combined Score Consistency Attempt pass rate Flaky tests Tests Correct Response Time (avg) Output Tokens Reasoning Tokens
DeepSeek V3.2 6.5 10.0 0.0% 0 115.89s 2,887 0
DeepSeek V4 Pro 9.5 10.0 100.0% 0 34.55s 826 0
Data parsing and extraction Score Consistency Attempt pass rate Flaky tests Tests Correct Response Time (avg) Output Tokens Reasoning Tokens
DeepSeek V3.2 6.3 5.8 66.7% 1 9.42s 1,710 0
DeepSeek V4 Pro 10.0 10.0 100.0% 0 54.04s 65 0
Domain specific Score Consistency Attempt pass rate Flaky tests Tests Correct Response Time (avg) Output Tokens Reasoning Tokens
DeepSeek V3.2 3.0 10.0 0.0% 0 1.52s 18 0
DeepSeek V4 Pro 5.3 10.0 33.3% 0 3.74s 6 0
General Intelligence Score Consistency Attempt pass rate Flaky tests Tests Correct Response Time (avg) Output Tokens Reasoning Tokens
DeepSeek V3.2 10.0 10.0 100.0% 0 2.86s 67 0
DeepSeek V4 Pro 4.5 10.0 0.0% 0 6.06s 45 0
Instructions following Score Consistency Attempt pass rate Flaky tests Tests Correct Response Time (avg) Output Tokens Reasoning Tokens
DeepSeek V3.2 10.0 10.0 100.0% 0 1.52s 66 0
DeepSeek V4 Pro 6.5 10.0 50.0% 0 3.57s 22 0
Puzzle Solving Score Consistency Attempt pass rate Flaky tests Tests Correct Response Time (avg) Output Tokens Reasoning Tokens
DeepSeek V3.2 8.5 7.5 88.9% 1 7.37s 1,136 0
DeepSeek V4 Pro 6.0 7.1 44.4% 1 28.25s 92 0
Tool Calling Score Consistency Attempt pass rate Flaky tests Tests Correct Response Time (avg) Output Tokens Reasoning Tokens
DeepSeek V3.2 10.0 10.0 100.0% 0 11.85s 522 0
DeepSeek V4 Pro 10.0 10.0 100.0% 0 6.47s 73 0

Quick Compare

Switch Comparison Pair