Compare · ModelsLive · 2 picked · head to head
DeepSeek V3.2 vs DeepSeek V4 Flash 0731
Side by side · benchmarks, pricing, and signals you can act on.
Winner summary
DeepSeek V4 Flash 0731 wins on 9/9 benchmarks
DeepSeek V4 Flash 0731 wins 9 of 9 shared benchmarks. Leads in reasoning · knowledge · general.
Category leads
reasoning·DeepSeek V4 Flash 0731knowledge·DeepSeek V4 Flash 0731general·DeepSeek V4 Flash 0731math·DeepSeek V4 Flash 0731
Hype vs Reality
Attention vs performance
DeepSeek V3.2
#142 by perf·no signal
DeepSeek V4 Flash 0731
#76 by perf·#6 by attention
Best value
DeepSeek V3.2
1.6x better value than DeepSeek V4 Flash 0731
DeepSeek V3.2
136.3 pts/$
$0.35/M
DeepSeek V4 Flash 0731
85.5 pts/$
$0.65/M
Vendor risk
Mixed exposure
One or more vendors flagged
DeepSeek
$3.4B·Tier 1
DeepSeek
$3.4B·Tier 1
Head to head
9 benchmarks · 2 models
DeepSeek V3.2DeepSeek V4 Flash 0731
ARC-AGI
DeepSeek V4 Flash 0731 leads by +32.0
ARC-AGI · the original Abstraction and Reasoning Corpus, testing whether AI can solve novel visual pattern recognition tasks without memorization.
DeepSeek V3.2
57.0
DeepSeek V4 Flash 0731
89.0
ARC-AGI-2
DeepSeek V4 Flash 0731 leads by +57.4
ARC-AGI-2 · the second iteration of the Abstraction and Reasoning Corpus, testing novel pattern recognition and abstract reasoning without prior training data.
DeepSeek V3.2
4.0
DeepSeek V4 Flash 0731
61.4
Chess Puzzles
DeepSeek V4 Flash 0731 leads by +20.0
Chess Puzzles · tests strategic and tactical reasoning by having models solve chess puzzle positions, evaluating lookahead and pattern recognition abilities.
DeepSeek V3.2
9.5
DeepSeek V4 Flash 0731
29.5
Dtbench
DeepSeek V4 Flash 0731 leads by +8.9
DeepSeek V3.2
76.0
DeepSeek V4 Flash 0731
84.9
GPQA diamond
DeepSeek V4 Flash 0731 leads by +10.1
Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs.
DeepSeek V3.2
77.9
DeepSeek V4 Flash 0731
88.0
Lmca
DeepSeek V4 Flash 0731 leads by +15.2
DeepSeek V3.2
33.9
DeepSeek V4 Flash 0731
49.1
OTIS Mock AIME 2024-2025
DeepSeek V4 Flash 0731 leads by +6.6
OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills.
DeepSeek V3.2
87.8
DeepSeek V4 Flash 0731
94.4
Proofbench
DeepSeek V4 Flash 0731 leads by +48.0
DeepSeek V3.2
8.0
DeepSeek V4 Flash 0731
56.0
SimpleQA Verified
DeepSeek V4 Flash 0731 leads by +6.1
SimpleQA Verified · short factual questions with verified answers, measuring factual accuracy and the tendency to hallucinate or provide incorrect information.
DeepSeek V3.2
27.5
DeepSeek V4 Flash 0731
33.6
Full benchmark table
| Benchmark | DeepSeek V3.2 | DeepSeek V4 Flash 0731 |
|---|---|---|
ARC-AGI ARC-AGI · the original Abstraction and Reasoning Corpus, testing whether AI can solve novel visual pattern recognition tasks without memorization. | 57.0 | 89.0 |
ARC-AGI-2 ARC-AGI-2 · the second iteration of the Abstraction and Reasoning Corpus, testing novel pattern recognition and abstract reasoning without prior training data. | 4.0 | 61.4 |
Chess Puzzles Chess Puzzles · tests strategic and tactical reasoning by having models solve chess puzzle positions, evaluating lookahead and pattern recognition abilities. | 9.5 | 29.5 |
Dtbench | 76.0 | 84.9 |
GPQA diamond Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs. | 77.9 | 88.0 |
Lmca | 33.9 | 49.1 |
OTIS Mock AIME 2024-2025 OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills. | 87.8 | 94.4 |
Proofbench | 8.0 | 56.0 |
SimpleQA Verified SimpleQA Verified · short factual questions with verified answers, measuring factual accuracy and the tendency to hallucinate or provide incorrect information. | 27.5 | 33.6 |
Pricing · per 1M tokens · projected $/mo at 10M tokens
| Model | Input | Output | Context | Projected $/mo |
|---|---|---|---|---|
| $0.28 | $0.42 | 164K tokens (~82 books) | $3.15 | |
| $0.02 | $1.28 | 1.0M tokens (~524 books) | $3.31 |