Compare · ModelsLive · 2 picked · head to head
DeepSeek V3 vs GPT-5.1
Side by side · benchmarks, pricing, and signals you can act on.
Winner summary
GPT-5.1 wins on 11/13 benchmarks
GPT-5.1 wins 11 of 13 shared benchmarks. Leads in arena · general · math.
Category leads
arena·GPT-5.1general·GPT-5.1math·GPT-5.1knowledge·DeepSeek V3language·GPT-5.1reasoning·GPT-5.1coding·GPT-5.1
Hype vs Reality
Attention vs performance
DeepSeek V3
#70 by perf·no signal
GPT-5.1
#138 by perf·#4 by attention
Best value
DeepSeek V3
10.2x better value than GPT-5.1
DeepSeek V3
87.2 pts/$
$0.64/M
GPT-5.1
8.6 pts/$
$5.63/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 V3GPT-5.1
Chatbot Arena Elo · Overall
GPT-5.1 leads by +80.1
DeepSeek V3
1358.4
GPT-5.1
1438.5
Dtbench
GPT-5.1 leads by +42.2
DeepSeek V3
41.3
GPT-5.1
83.5
FrontierMath-2025-02-28-Private
GPT-5.1 leads by +51.4
FrontierMath (Feb 2025) · original research-level math problems created by mathematicians, testing capabilities at the boundary of current AI mathematical reasoning.
DeepSeek V3
3.0
GPT-5.1
54.5
GPQA diamond
GPT-5.1 leads by +41.5
Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs.
DeepSeek V3
42.0
GPT-5.1
83.5
HELM · GPQA
DeepSeek V3 leads by +9.6
DeepSeek V3
53.8
GPT-5.1
44.2
HELM · IFEval
GPT-5.1 leads by +10.3
DeepSeek V3
83.2
GPT-5.1
93.5
HELM · MMLU-Pro
DeepSeek V3 leads by +14.4
DeepSeek V3
72.3
GPT-5.1
57.9
HELM · Omni-MATH
GPT-5.1 leads by +6.1
DeepSeek V3
40.3
GPT-5.1
46.4
HELM · WildBench
GPT-5.1 leads by +3.2
DeepSeek V3
83.1
GPT-5.1
86.3
Lmca
GPT-5.1 leads by +33.4
DeepSeek V3
18.2
GPT-5.1
51.6
OTIS Mock AIME 2024-2025
GPT-5.1 leads by +72.8
OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills.
DeepSeek V3
15.8
GPT-5.1
88.6
SimpleBench
GPT-5.1 leads by +41.2
SimpleBench · tests fundamental reasoning capabilities with straightforward problems designed to expose gaps in basic logical and spatial thinking.
DeepSeek V3
2.7
GPT-5.1
43.8
WeirdML
GPT-5.1 leads by +24.7
WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns.
DeepSeek V3
36.1
GPT-5.1
60.8
Full benchmark table
| Benchmark | DeepSeek V3 | GPT-5.1 |
|---|---|---|
Chatbot Arena Elo · Overall | 1358.4 | 1438.5 |
Dtbench | 41.3 | 83.5 |
FrontierMath-2025-02-28-Private FrontierMath (Feb 2025) · original research-level math problems created by mathematicians, testing capabilities at the boundary of current AI mathematical reasoning. | 3.0 | 54.5 |
GPQA diamond Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs. | 42.0 | 83.5 |
HELM · GPQA | 53.8 | 44.2 |
HELM · IFEval | 83.2 | 93.5 |
HELM · MMLU-Pro | 72.3 | 57.9 |
HELM · Omni-MATH | 40.3 | 46.4 |
HELM · WildBench | 83.1 | 86.3 |
Lmca | 18.2 | 51.6 |
OTIS Mock AIME 2024-2025 OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills. | 15.8 | 88.6 |
SimpleBench SimpleBench · tests fundamental reasoning capabilities with straightforward problems designed to expose gaps in basic logical and spatial thinking. | 2.7 | 43.8 |
WeirdML WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns. | 36.1 | 60.8 |
Pricing · per 1M tokens · projected $/mo at 10M tokens
| Model | Input | Output | Context | Projected $/mo |
|---|---|---|---|---|
| $0.26 | $1.03 | 164K tokens (~82 books) | $4.50 | |
| $1.25 | $10.00 | 400K tokens (~200 books) | $34.38 |
People also compared