Compare · ModelsLive · 2 picked · head to head
Gemini 3 Pro vs GPT-5.1
Side by side · benchmarks, pricing, and signals you can act on.
Winner summary
Gemini 3 Pro wins on 21/25 benchmarks
Gemini 3 Pro wins 21 of 25 shared benchmarks. Leads in agentic · reasoning · arena.
Category leads
agentic·Gemini 3 Proreasoning·Gemini 3 Proarena·Gemini 3 Proknowledge·Gemini 3 Progeneral·GPT-5.1math·Gemini 3 Procoding·Gemini 3 Prolanguage·GPT-5.1
Head to head · from the data · data as of 2026-11-17
- Benchmarks · Gemini 3 Pro leads on 21 of 25 shared benchmarks (APEX-Agents, ARC-AGI, ARC-AGI-2, Chatbot Arena Elo · Coding, 17 more), GPT-5.1 on 4 (Chess Puzzles, Cl Bench, HELM · IFEval, HELM · WildBench)
- Cheapest provider · GPT-5.1 at OpenAI $0.63/M input (of 5 offers) as of 2026-11-17
- Price history · GPT-5.1 OpenRouter input list price had no move of 5% or more in the 90 days to 2026-11-17
Sources · scores from public leaderboards, list prices and provider offers from OpenRouter, price moves from the BenchGecko changelog (5% or more, held in the next daily check), Gecko tests run by BenchGecko.
Hype vs Reality
Attention vs performance
Gemini 3 Pro
#78 by perf·#9 by attention
GPT-5.1
#157 by perf·#11 by attention
Vendor risk
Who is behind the model
Google DeepMind
$3.49T·Tier 1
OpenAI
$840.0B·Tier 1
Head to head
25 benchmarks · 2 models
Gemini 3 ProGPT-5.1
APEX-Agents
Gemini 3 Pro leads by +0.9
APEX-Agents · evaluates AI agents on complex, multi-step tasks requiring planning, tool use, and autonomous decision-making in realistic environments.
Gemini 3 Pro
18.4
GPT-5.1
17.5
ARC-AGI
Gemini 3 Pro leads by +2.2
ARC-AGI · the original Abstraction and Reasoning Corpus, testing whether AI can solve novel visual pattern recognition tasks without memorization.
Gemini 3 Pro
75.0
GPT-5.1
72.8
ARC-AGI-2
Gemini 3 Pro leads by +13.5
ARC-AGI-2 · the second iteration of the Abstraction and Reasoning Corpus, testing novel pattern recognition and abstract reasoning without prior training data.
Gemini 3 Pro
31.1
GPT-5.1
17.6
Chatbot Arena Elo · Coding
Gemini 3 Pro leads by +97.9
Gemini 3 Pro
1439.0
GPT-5.1
1341.0
Chatbot Arena Elo · Overall
Gemini 3 Pro leads by +47.0
Gemini 3 Pro
1485.5
GPT-5.1
1438.5
Chess Puzzles
GPT-5.1 leads by +11.6
Chess Puzzles · tests strategic and tactical reasoning by having models solve chess puzzle positions, evaluating lookahead and pattern recognition abilities.
Gemini 3 Pro
27.4
GPT-5.1
39.0
Cl Bench
GPT-5.1 leads by +7.9
Gemini 3 Pro
15.8
GPT-5.1
23.7
DeepResearch Bench
Gemini 3 Pro leads by +3.5
DeepResearch Bench · evaluates AI on complex multi-step research tasks requiring information gathering, synthesis, and producing comprehensive analyses.
Gemini 3 Pro
46.3
GPT-5.1
42.8
FrontierMath-2025-02-28-Private
Gemini 3 Pro leads by +34.9
FrontierMath (Feb 2025) · original research-level math problems created by mathematicians, testing capabilities at the boundary of current AI mathematical reasoning.
Gemini 3 Pro
66.0
GPT-5.1
31.0
FrontierMath-Tier-4-2025-07-01-Private
Gemini 3 Pro leads by +18.8
FrontierMath Tier 4 (Jul 2025) · the most challenging tier of frontier mathematics, containing problems that push the absolute limits of AI mathematical reasoning.
Gemini 3 Pro
31.3
GPT-5.1
12.5
GPQA diamond
Gemini 3 Pro leads by +6.9
Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs.
Gemini 3 Pro
93.9
GPT-5.1
87.0
GSO-Bench
Gemini 3 Pro leads by +4.9
GSO-Bench · evaluates AI models on real-world open-source software engineering tasks, testing the ability to understand and resolve actual GitHub issues.
Gemini 3 Pro
18.6
GPT-5.1
13.7
HELM · GPQA
Gemini 3 Pro leads by +36.1
Gemini 3 Pro
80.3
GPT-5.1
44.2
HELM · IFEval
GPT-5.1 leads by +5.9
Gemini 3 Pro
87.6
GPT-5.1
93.5
HELM · MMLU-Pro
Gemini 3 Pro leads by +32.4
Gemini 3 Pro
90.3
GPT-5.1
57.9
HELM · Omni-MATH
Gemini 3 Pro leads by +9.2
Gemini 3 Pro
55.6
GPT-5.1
46.4
HELM · WildBench
GPT-5.1 leads by +0.4
Gemini 3 Pro
85.9
GPT-5.1
86.3
HLE
Gemini 3 Pro leads by +14.5
HLE (Humanity's Last Exam) · a reasoning benchmark designed to be the hardest public evaluation of AI. Questions span mathematics, physics, philosophy, and logic · curated to be at or beyond the frontier of human expert capability. Tested with and without tool augmentation.
Gemini 3 Pro
34.4
GPT-5.1
19.8
OTIS Mock AIME 2024-2025
Gemini 3 Pro leads by +2.8
OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills.
Gemini 3 Pro
91.4
GPT-5.1
88.6
SimpleBench
Gemini 3 Pro leads by +27.8
SimpleBench · tests fundamental reasoning capabilities with straightforward problems designed to expose gaps in basic logical and spatial thinking.
Gemini 3 Pro
71.7
GPT-5.1
43.8
SimpleQA Verified
Gemini 3 Pro leads by +24.9
SimpleQA Verified · short factual questions with verified answers, measuring factual accuracy and the tendency to hallucinate or provide incorrect information.
Gemini 3 Pro
72.9
GPT-5.1
48.0
SWE-Bench verified
Gemini 3 Pro leads by +5.0
SWE-bench Verified · 500 human-validated tasks from 12 real Python repositories (Django, Flask, scikit-learn, sympy, and others). Each task requires the model to produce a git patch that resolves a real GitHub issue and passes the test suite. The verified subset eliminates ambiguous tasks from the original SWE-bench. The benchmark remains the most-cited evaluation for code-generation capability.
Gemini 3 Pro
72.9
GPT-5.1
68.0
Terminal Bench
Gemini 3 Pro leads by +21.8
Terminal-Bench 2.0 · evaluates AI agents on real terminal-based coding tasks · writing scripts, debugging, running tests, and managing projects entirely through command-line interaction. Tests both code quality and terminal fluency.
Gemini 3 Pro
69.4
GPT-5.1
47.6
VPCT
Gemini 3 Pro leads by +48.5
VPCT (Visual Pattern Completion Test) · tests visual reasoning and pattern recognition by having models complete visual sequences and transformations.
Gemini 3 Pro
86.5
GPT-5.1
38.0
WeirdML
Gemini 3 Pro leads by +9.2
WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns.
Gemini 3 Pro
69.9
GPT-5.1
60.8
Full benchmark table
| Benchmark | Gemini 3 Pro | GPT-5.1 |
|---|---|---|
APEX-Agents APEX-Agents · evaluates AI agents on complex, multi-step tasks requiring planning, tool use, and autonomous decision-making in realistic environments. | 18.4 | 17.5 |
ARC-AGI ARC-AGI · the original Abstraction and Reasoning Corpus, testing whether AI can solve novel visual pattern recognition tasks without memorization. | 75.0 | 72.8 |
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. | 31.1 | 17.6 |
Chatbot Arena Elo · Coding | 1439.0 | 1341.0 |
Chatbot Arena Elo · Overall | 1485.5 | 1438.5 |
Chess Puzzles Chess Puzzles · tests strategic and tactical reasoning by having models solve chess puzzle positions, evaluating lookahead and pattern recognition abilities. | 27.4 | 39.0 |
Cl Bench | 15.8 | 23.7 |
DeepResearch Bench DeepResearch Bench · evaluates AI on complex multi-step research tasks requiring information gathering, synthesis, and producing comprehensive analyses. | 46.3 | 42.8 |
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. | 66.0 | 31.0 |
FrontierMath-Tier-4-2025-07-01-Private FrontierMath Tier 4 (Jul 2025) · the most challenging tier of frontier mathematics, containing problems that push the absolute limits of AI mathematical reasoning. | 31.3 | 12.5 |
GPQA diamond Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs. | 93.9 | 87.0 |
GSO-Bench GSO-Bench · evaluates AI models on real-world open-source software engineering tasks, testing the ability to understand and resolve actual GitHub issues. | 18.6 | 13.7 |
HELM · GPQA | 80.3 | 44.2 |
HELM · IFEval | 87.6 | 93.5 |
HELM · MMLU-Pro | 90.3 | 57.9 |
HELM · Omni-MATH | 55.6 | 46.4 |
HELM · WildBench | 85.9 | 86.3 |
HLE HLE (Humanity's Last Exam) · a reasoning benchmark designed to be the hardest public evaluation of AI. Questions span mathematics, physics, philosophy, and logic · curated to be at or beyond the frontier of human expert capability. Tested with and without tool augmentation. | 34.4 | 19.8 |
OTIS Mock AIME 2024-2025 OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills. | 91.4 | 88.6 |
SimpleBench SimpleBench · tests fundamental reasoning capabilities with straightforward problems designed to expose gaps in basic logical and spatial thinking. | 71.7 | 43.8 |
SimpleQA Verified SimpleQA Verified · short factual questions with verified answers, measuring factual accuracy and the tendency to hallucinate or provide incorrect information. | 72.9 | 48.0 |
SWE-Bench verified SWE-bench Verified · 500 human-validated tasks from 12 real Python repositories (Django, Flask, scikit-learn, sympy, and others). Each task requires the model to produce a git patch that resolves a real GitHub issue and passes the test suite. The verified subset eliminates ambiguous tasks from the original SWE-bench. The benchmark remains the most-cited evaluation for code-generation capability. | 72.9 | 68.0 |
Terminal Bench Terminal-Bench 2.0 · evaluates AI agents on real terminal-based coding tasks · writing scripts, debugging, running tests, and managing projects entirely through command-line interaction. Tests both code quality and terminal fluency. | 69.4 | 47.6 |
VPCT VPCT (Visual Pattern Completion Test) · tests visual reasoning and pattern recognition by having models complete visual sequences and transformations. | 86.5 | 38.0 |
WeirdML WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns. | 69.9 | 60.8 |
Pricing · per 1M tokens · projected $/mo at 10M tokens
| Model | Input | Output | Context | Projected $/mo |
|---|---|---|---|---|
| n/a | n/a | n/a | n/a | |
| $1.25 | $10.00 | 400K tokens (~200 books) | $34.38 |
Gemini 3 Pro vs GPT-5.1 · questions answered from the data
Gemini 3 Pro leads on 21 of 25 shared benchmarks (APEX-Agents, ARC-AGI, ARC-AGI-2, Chatbot Arena Elo · Coding, 17 more), GPT-5.1 on 4 (Chess Puzzles, Cl Bench, HELM · IFEval, HELM · WildBench)
People also compared