Compare · ModelsLive · 2 picked · head to head
DeepSeek V4 Flash 0731 vs Qwen3.8 Max (0902)
Side by side · benchmarks, pricing, and signals you can act on.
Winner summary
Qwen3.8 Max (0902) wins on 10/10 benchmarks
Qwen3.8 Max (0902) wins 10 of 10 shared benchmarks. Leads in knowledge · general · math.
Category leads
knowledge·Qwen3.8 Max (0902)general·Qwen3.8 Max (0902)math·Qwen3.8 Max (0902)
Hype vs Reality
Attention vs performance
DeepSeek V4 Flash 0731
#80 by perf·#10 by attention
Qwen3.8 Max (0902)
#85 by perf·#2 by attention
Best value
DeepSeek V4 Flash 0731
12.3x better value than Qwen3.8 Max (0902)
DeepSeek V4 Flash 0731
168.6 pts/$
$0.33/M
Qwen3.8 Max (0902)
13.7 pts/$
$4.00/M
Vendor risk
Mixed exposure
One or more vendors flagged
DeepSeek
$3.4B·Tier 1
Alibaba (Qwen)
$293.0B·Tier 1
Head to head
10 benchmarks · 2 models
DeepSeek V4 Flash 0731Qwen3.8 Max (0902)
Chess Puzzles
Qwen3.8 Max (0902) leads by +14.7
Chess Puzzles · tests strategic and tactical reasoning by having models solve chess puzzle positions, evaluating lookahead and pattern recognition abilities.
DeepSeek V4 Flash 0731
28.4
Qwen3.8 Max (0902)
43.2
Dtbench
Qwen3.8 Max (0902) leads by +1.8
DeepSeek V4 Flash 0731
84.9
Qwen3.8 Max (0902)
86.7
FrontierMath-Tier-4-v2-Private
Qwen3.8 Max (0902) leads by +22.0
DeepSeek V4 Flash 0731
24.4
Qwen3.8 Max (0902)
46.3
FrontierMath-Tiers-1-3-v2-Private
Qwen3.8 Max (0902) leads by +17.2
DeepSeek V4 Flash 0731
57.5
Qwen3.8 Max (0902)
74.7
GPQA diamond
Qwen3.8 Max (0902) leads by +2.2
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
Qwen3.8 Max (0902)
90.2
Lmca
Qwen3.8 Max (0902) leads by +5.3
DeepSeek V4 Flash 0731
49.1
Qwen3.8 Max (0902)
54.4
Mystery Game Puzzles
Qwen3.8 Max (0902) leads by +4.4
DeepSeek V4 Flash 0731
27.3
Qwen3.8 Max (0902)
31.7
OTIS Mock AIME 2024-2025
Qwen3.8 Max (0902) leads by +5.0
OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills.
DeepSeek V4 Flash 0731
94.4
Qwen3.8 Max (0902)
99.4
Proofbench
Qwen3.8 Max (0902) leads by +2.0
DeepSeek V4 Flash 0731
56.0
Qwen3.8 Max (0902)
58.0
SimpleQA Verified
Qwen3.8 Max (0902) leads by +12.2
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
Qwen3.8 Max (0902)
45.8
Full benchmark table
| Benchmark | DeepSeek V4 Flash 0731 | Qwen3.8 Max (0902) |
|---|---|---|
Chess Puzzles Chess Puzzles · tests strategic and tactical reasoning by having models solve chess puzzle positions, evaluating lookahead and pattern recognition abilities. | 28.4 | 43.2 |
Dtbench | 84.9 | 86.7 |
FrontierMath-Tier-4-v2-Private | 24.4 | 46.3 |
FrontierMath-Tiers-1-3-v2-Private | 57.5 | 74.7 |
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 | 90.2 |
Lmca | 49.1 | 54.4 |
Mystery Game Puzzles | 27.3 | 31.7 |
OTIS Mock AIME 2024-2025 OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills. | 94.4 | 99.4 |
Proofbench | 56.0 | 58.0 |
SimpleQA Verified SimpleQA Verified · short factual questions with verified answers, measuring factual accuracy and the tendency to hallucinate or provide incorrect information. | 33.6 | 45.8 |
Pricing · per 1M tokens · projected $/mo at 10M tokens
| Model | Input | Output | Context | Projected $/mo |
|---|---|---|---|---|
| $0.02 | $0.64 | 1.0M tokens (~524 books) | $1.72 | |
| $2.00 | $6.00 | 1.0M tokens (~500 books) | $30.00 |