Compare · ModelsLive · 2 picked · head to head
DeepSeek V4 Flash 0731 vs GPT-5 Mini
Side by side · benchmarks, pricing, and signals you can act on.
Winner summary
DeepSeek V4 Flash 0731 wins on 13/13 benchmarks
DeepSeek V4 Flash 0731 wins 13 of 13 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 0731coding·DeepSeek V4 Flash 0731
Hype vs Reality
Attention vs performance
DeepSeek V4 Flash 0731
#76 by perf·#6 by attention
GPT-5 Mini
#108 by perf·#4 by attention
Best value
DeepSeek V4 Flash 0731
1.9x better value than GPT-5 Mini
DeepSeek V4 Flash 0731
85.5 pts/$
$0.65/M
GPT-5 Mini
45.8 pts/$
$1.13/M
Vendor risk
Mixed exposure
One or more vendors flagged
DeepSeek
$3.4B·Tier 1
OpenAI
$840.0B·Tier 1
Head to head
13 benchmarks · 2 models
DeepSeek V4 Flash 0731GPT-5 Mini
ARC-AGI
DeepSeek V4 Flash 0731 leads by +34.7
ARC-AGI · the original Abstraction and Reasoning Corpus, testing whether AI can solve novel visual pattern recognition tasks without memorization.
DeepSeek V4 Flash 0731
89.0
GPT-5 Mini
54.3
ARC-AGI-2
DeepSeek V4 Flash 0731 leads by +57.0
ARC-AGI-2 · the second iteration of the Abstraction and Reasoning Corpus, testing novel pattern recognition and abstract reasoning without prior training data.
DeepSeek V4 Flash 0731
61.4
GPT-5 Mini
4.4
Chess Puzzles
DeepSeek V4 Flash 0731 leads by +3.1
Chess Puzzles · tests strategic and tactical reasoning by having models solve chess puzzle positions, evaluating lookahead and pattern recognition abilities.
DeepSeek V4 Flash 0731
29.5
GPT-5 Mini
26.4
Dtbench
DeepSeek V4 Flash 0731 leads by +17.3
DeepSeek V4 Flash 0731
84.9
GPT-5 Mini
67.5
FrontierMath-Tier-4-v2-Private
DeepSeek V4 Flash 0731 leads by +12.2
DeepSeek V4 Flash 0731
24.4
GPT-5 Mini
12.2
FrontierMath-Tiers-1-3-v2-Private
DeepSeek V4 Flash 0731 leads by +10.9
DeepSeek V4 Flash 0731
57.5
GPT-5 Mini
46.7
GPQA diamond
DeepSeek V4 Flash 0731 leads by +21.4
Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs.
DeepSeek V4 Flash 0731
88.0
GPT-5 Mini
66.7
Lmca
DeepSeek V4 Flash 0731 leads by +8.9
DeepSeek V4 Flash 0731
49.1
GPT-5 Mini
40.2
Mystery Game Puzzles
DeepSeek V4 Flash 0731 leads by +26.4
DeepSeek V4 Flash 0731
27.3
GPT-5 Mini
0.9
OTIS Mock AIME 2024-2025
DeepSeek V4 Flash 0731 leads by +7.8
OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills.
DeepSeek V4 Flash 0731
94.4
GPT-5 Mini
86.7
Proofbench
DeepSeek V4 Flash 0731 leads by +47.0
DeepSeek V4 Flash 0731
56.0
GPT-5 Mini
9.0
SimpleQA Verified
DeepSeek V4 Flash 0731 leads by +12.0
SimpleQA Verified · short factual questions with verified answers, measuring factual accuracy and the tendency to hallucinate or provide incorrect information.
DeepSeek V4 Flash 0731
33.6
GPT-5 Mini
21.6
WeirdML
DeepSeek V4 Flash 0731 leads by +10.3
WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns.
DeepSeek V4 Flash 0731
63.0
GPT-5 Mini
52.7
Full benchmark table
| Benchmark | DeepSeek V4 Flash 0731 | GPT-5 Mini |
|---|---|---|
ARC-AGI ARC-AGI · the original Abstraction and Reasoning Corpus, testing whether AI can solve novel visual pattern recognition tasks without memorization. | 89.0 | 54.3 |
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. | 61.4 | 4.4 |
Chess Puzzles Chess Puzzles · tests strategic and tactical reasoning by having models solve chess puzzle positions, evaluating lookahead and pattern recognition abilities. | 29.5 | 26.4 |
Dtbench | 84.9 | 67.5 |
FrontierMath-Tier-4-v2-Private | 24.4 | 12.2 |
FrontierMath-Tiers-1-3-v2-Private | 57.5 | 46.7 |
GPQA diamond Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs. | 88.0 | 66.7 |
Lmca | 49.1 | 40.2 |
Mystery Game Puzzles | 27.3 | 0.9 |
OTIS Mock AIME 2024-2025 OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills. | 94.4 | 86.7 |
Proofbench | 56.0 | 9.0 |
SimpleQA Verified SimpleQA Verified · short factual questions with verified answers, measuring factual accuracy and the tendency to hallucinate or provide incorrect information. | 33.6 | 21.6 |
WeirdML WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns. | 63.0 | 52.7 |
Pricing · per 1M tokens · projected $/mo at 10M tokens
| Model | Input | Output | Context | Projected $/mo |
|---|---|---|---|---|
| $0.02 | $1.28 | 1.0M tokens (~524 books) | $3.31 | |
| $0.25 | $2.00 | 400K tokens (~200 books) | $6.88 |