Model comparison
Gemini 3 Pro vs GPT-4o
Gemini 3 Pro is the stronger model overall, scoring 54.8 to 28.6 on the Noometry Index.
Last verified . 46 shared benchmarks.
GoogleSummary
- They share 46 benchmarks with published results for both. Gemini 3 Pro scores higher in 10 categories and GPT-4o in 0 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Gemini 3 Pro leads 52.5 to 9.4.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 91.4% for Gemini 3 Pro and 6.4% for GPT-4o.
Side by side
| Gemini 3 Pro | GPT-4o | |
|---|---|---|
| Provider | OpenAI | |
| Noometry Index | 54.8 | 28.6 |
| Released | 2025-11-18 | 2024-05-13 |
| Weights | Proprietary | Proprietary |
| Context window | — | 128K |
| Max output | — | 16K |
| Input $ / M tokens | — | $2.50 |
| Output $ / M tokens | — | $10 |
| Results tracked | 67 | 72 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Gemini 3 Pro leads
Gemini 3 Pro: 51.6 (#39), GPT-4o: 24.8 (#328)
| Benchmark | Gemini 3 Pro | GPT-4o |
|---|---|---|
| SWE-bench Verified | 72.9% | 31% |
| SWE-bench Verified (bash only) | 74.2% | 21.6% |
| GSO | 18.6% | 0% |
| WeirdML | 69.9% | 25.1% |
| LMArena Coding | 1481 | 1297 |
| Aider Polyglot | — | 45.3% |
| LMArena WebDev | 1440 | — |
| SWE-bench Multilingual | 68.7% | — |
| BigCodeBench Instruct | — | 51.1% |
| LiveBench Coding | — | 51.4% |
| BigCodeBench Complete | — | 61.1% |
| CadEval | — | 26% |
| ALE-Bench | 1,177 | — |
| AlgoTune | 1.83 | — |
| HumanEval+ | — | 87.2% |
| MBPP+ | — | 72.2% |
Agentic & Tool Use Gemini 3 Pro leads
Gemini 3 Pro: 40.6 (#23), GPT-4o: 21.0 (#141)
| Benchmark | Gemini 3 Pro | GPT-4o |
|---|---|---|
| GDPval | 40.3% | 9.9% |
| BALROG | 58.1% | 32.3% |
| LMArena Search | 1207 | 1006 |
| METR Time Horizons | 71% | 40.8% |
| Terminal-Bench | 69.4% | — |
| Berkeley Function Calling Leaderboard | 72.5% | — |
| Remote Labor Index | 1.3% | — |
| TheAgentCompany | — | 8.6% |
| τ²-bench Airline | 80.5% | — |
| τ²-bench Banking | 18% | — |
| τ²-bench Retail | 75.9% | — |
| τ²-bench Telecom | 91% | — |
| Cybench | — | 12.5% |
| DeepResearch Bench | 46.3% | — |
| Vending-Bench 2 | 5,478 | — |
Reasoning Gemini 3 Pro leads
Gemini 3 Pro: 52.5 (#31), GPT-4o: 9.4 (#343)
| Benchmark | Gemini 3 Pro | GPT-4o |
|---|---|---|
| ARC-AGI-2 | 31.1% | 0% |
| SimpleBench | 76.4% | 17.8% |
| ARC-AGI-1 | 75% | 4.5% |
| CritPt | 6.9% | 0% |
| Chess Puzzles | 31% | 13% |
| EnigmaEval | 18.2% | 0.8% |
| LMArena Hard Prompts | 1480 | 1281 |
| Epoch Capabilities Index | 152.92 | 128.97 |
| ForecastBench | 61.2 | 57.7 |
| Kagi LLM Benchmark | 80.1% | — |
| NYT Connections (extended) | 94.4% | — |
| LiveBench Reasoning | — | 55.8% |
| DTBench | — | 64.5% |
| LiveBench Data Analysis | — | 60.9% |
| LMCA | — | 16.6% |
| LiveBench | — | 55.3% |
Math Gemini 3 Pro leads
Gemini 3 Pro: 49.9 (#59), GPT-4o: 10.6 (#312)
| Benchmark | Gemini 3 Pro | GPT-4o |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 91.4% | 6.4% |
| Omni-MATH | 55.5% | 29.3% |
| LMArena Math | 1476 | 1285 |
| FrontierMath (Feb 2025 set) | 37.6% | 0.3% |
| FrontierMath (Tiers 1-3) | — | 0.4% |
| MathArena Final-Answer Competitions | 67% | — |
| ProofBench | 20% | — |
| LiveBench Math | — | 49.5% |
| MATH Level 5 | — | 53.3% |
| FrontierMath Tier 4 (v1) | 18.8% | — |
Knowledge Gemini 3 Pro leads
Gemini 3 Pro: 64.4 (#16), GPT-4o: 28.8 (#242)
| Benchmark | Gemini 3 Pro | GPT-4o |
|---|---|---|
| GPQA Diamond | 92.6% | 49.2% |
| Humanity's Last Exam | 37.5% | 2.7% |
| MMLU-Pro | 90.3% | 71.3% |
| Vectara Hallucination Rate | 13.6% | 9.6% |
| GPQA (HELM) | 80.3% | 52% |
| LMArena Expert | 1475 | 1250 |
| SimpleQA Verified | — | 26% |
| Confabulations | — | 15.3% |
| MMLU | — | 88.1% |
Multimodal Gemini 3 Pro leads
Gemini 3 Pro: 57.6 (#2), GPT-4o: 34.5 (#91)
| Benchmark | Gemini 3 Pro | GPT-4o |
|---|---|---|
| LMArena Vision | 1305 | 1137 |
| GeoBench | 84% | 71% |
| VPCT | 91% | 40% |
| Video-MME | — | 71.9% |
| LMArena Document | 1434 | — |
| ScienceQA | — | 88.5% |
Multilingual Gemini 3 Pro leads
Gemini 3 Pro: 56.9 (#16), GPT-4o: 43.2 (#186)
| Benchmark | Gemini 3 Pro | GPT-4o |
|---|---|---|
| LMArena Non-English | 1474 | 1283 |
| LMArena Chinese | 1523 | 1277 |
| LMArena French | 1492 | 1304 |
| LMArena German | 1515 | 1282 |
| LMArena Japanese | 1510 | 1257 |
| LMArena Korean | 1448 | 1234 |
| LMArena Russian | 1493 | 1286 |
| LMArena Spanish | 1470 | 1292 |
Instruction Following Gemini 3 Pro leads
Gemini 3 Pro: 76.3 (#45), GPT-4o: 66.6 (#207)
| Benchmark | Gemini 3 Pro | GPT-4o |
|---|---|---|
| IFEval | 87.7% | 81.7% |
| LMArena Instruction Following | 1458 | 1278 |
| LiveBench Instruction Following | — | 68.6% |
Long Context Gemini 3 Pro leads
Gemini 3 Pro: 44.0 (#79), GPT-4o: 39.4 (#179)
| Benchmark | Gemini 3 Pro | GPT-4o |
|---|---|---|
| LMArena Longer Query | 1471 | 1289 |
| Fiction.LiveBench | — | 66.7% |
| CL-bench | 15.8% | — |
Writing & Preference Gemini 3 Pro leads
Gemini 3 Pro: 66.4 (#35), GPT-4o: 52.6 (#166)
| Benchmark | Gemini 3 Pro | GPT-4o |
|---|---|---|
| LMArena Text | 1479 | 1300 |
| LMArena Creative Writing | 1482 | 1292 |
| WildBench | 85.9% | 82.8% |
| LMArena Multi-Turn | 1484 | 1302 |
| Short-Story Creative Writing | — | 81.8% |
| EQ-Bench Creative Writing | 1525 | — |
| LiveBench Language | — | 47.6% |
Frequently asked questions
Is Gemini 3 Pro better than GPT-4o?
Gemini 3 Pro is the stronger model overall, scoring 54.8 to 28.6 on the Noometry Index.
Is Gemini 3 Pro or GPT-4o better for coding?
Gemini 3 Pro scores higher on coding benchmarks: 51.6 versus 24.8 in the Noometry coding category.
How many benchmarks do Gemini 3 Pro and GPT-4o share?
46 benchmarks have published results for both models. Gemini 3 Pro has 67 scored results on Noometry and GPT-4o has 72.