Model comparison
Gemini 3 Pro vs GPT-5.5
GPT-5.5 is the stronger model overall, scoring 63.4 to 54.8 on the Noometry Index.
Last verified . 47 shared benchmarks.
GoogleSummary
- They share 47 benchmarks with published results for both. Gemini 3 Pro scores higher in 3 categories and GPT-5.5 in 7 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-5.5 leads 81.7 to 49.9.
- The biggest single-benchmark swing is ARC-AGI-2: 31.1% for Gemini 3 Pro and 85% for GPT-5.5.
Side by side
| Gemini 3 Pro | GPT-5.5 | |
|---|---|---|
| Provider | OpenAI | |
| Noometry Index | 54.8 | 63.4 |
| Released | 2025-11-18 | 2026-04-23 |
| Weights | Proprietary | Proprietary |
| Context window | — | 1.05M |
| Max output | — | 128K |
| Input $ / M tokens | — | $5 |
| Output $ / M tokens | — | $30 |
| Results tracked | 67 | 71 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5.5 leads
Gemini 3 Pro: 51.6 (#39), GPT-5.5: 58.2 (#17)
| Benchmark | Gemini 3 Pro | GPT-5.5 |
|---|---|---|
| SWE-bench Verified | 72.9% | 80.6% |
| LMArena WebDev | 1440 | 1513 |
| GSO | 18.6% | 40.2% |
| WeirdML | 69.9% | 84.9% |
| LMArena Coding | 1481 | 1494 |
| ALE-Bench | 1,177 | 1,943 |
| DeepSWE | — | 67% |
| FrontierCode | — | 43% |
| SWE-bench Verified (bash only) | 74.2% | — |
| SWE-bench Multilingual | 68.7% | — |
| SciCode | — | 56.1% |
| MirrorCode | — | 10% |
| AlgoTune | 1.83 | — |
Agentic & Tool Use GPT-5.5 leads
Gemini 3 Pro: 40.6 (#23), GPT-5.5: 50.7 (#6)
| Benchmark | Gemini 3 Pro | GPT-5.5 |
|---|---|---|
| Terminal-Bench | 69.4% | 84.7% |
| Remote Labor Index | 1.3% | 6.3% |
| τ²-bench Banking | 18% | 44.6% |
| DeepResearch Bench | 46.3% | 54% |
| LMArena Search | 1207 | 1242 |
| Vending-Bench 2 | 5,478 | 7,524 |
| APEX-Agents | — | 55.1% |
| Berkeley Function Calling Leaderboard | 72.5% | — |
| OSWorld 2.0 | — | 13% |
| GDPval | 40.3% | — |
| τ²-bench Airline | 80.5% | — |
| τ²-bench Retail | 75.9% | — |
| τ²-bench Telecom | 91% | — |
| PostTrainBench | — | 27.2% |
| BALROG | 58.1% | — |
| ExploitBench | — | 47.4% |
| GBAEval | — | 53.2% |
| GDP.pdf | — | 26% |
| METR Time Horizons | 71% | — |
Reasoning GPT-5.5 leads
Gemini 3 Pro: 52.5 (#31), GPT-5.5: 72.8 (#11)
| Benchmark | Gemini 3 Pro | GPT-5.5 |
|---|---|---|
| ARC-AGI-2 | 31.1% | 85% |
| SimpleBench | 76.4% | 69% |
| Kagi LLM Benchmark | 80.1% | 88.8% |
| NYT Connections (extended) | 94.4% | 96.2% |
| ARC-AGI-1 | 75% | 95% |
| CritPt | 6.9% | 27.1% |
| Chess Puzzles | 31% | 54% |
| LMArena Hard Prompts | 1480 | 1489 |
| Epoch Capabilities Index | 152.92 | 159.1 |
| ForecastBench | 61.2 | 60.6 |
| EnigmaEval | 18.2% | — |
| EBR-Bench | — | 34.3% |
| Mystery Game Puzzles | — | 56% |
| DTBench | — | 96% |
| LMCA | — | 54.3% |
| Surface Evolver Bench | — | 88.1% |
| Bench to the Future 3 | — | 0.14 |
Math GPT-5.5 leads
Gemini 3 Pro: 49.9 (#59), GPT-5.5: 81.7 (#11)
| Benchmark | Gemini 3 Pro | GPT-5.5 |
|---|---|---|
| MathArena Final-Answer Competitions | 67% | 94.3% |
| OTIS Mock AIME 2024-2025 | 91.4% | 100% |
| ProofBench | 20% | 50% |
| LMArena Math | 1476 | 1486 |
| FrontierMath (Feb 2025 set) | 37.6% | 51.7% |
| FrontierMath Tier 4 (v1) | 18.8% | 35.4% |
| FrontierMath (Tiers 1-3) | — | 85.3% |
| FrontierMath Tier 4 | — | 72.5% |
| Omni-MATH | 55.5% | — |
| FrontierMath Erdős | — | 0% |
Knowledge Too close to call
Gemini 3 Pro: 64.4 (#16), GPT-5.5: 64.4 (#17)
| Benchmark | Gemini 3 Pro | GPT-5.5 |
|---|---|---|
| GPQA Diamond | 92.6% | 94% |
| Vectara Hallucination Rate | 13.6% | 9.3% |
| LMArena Expert | 1475 | 1508 |
| Humanity's Last Exam | 37.5% | — |
| SimpleQA Verified | — | 63% |
| MMLU-Pro | 90.3% | — |
| GPQA (HELM) | 80.3% | — |
Multimodal Gemini 3 Pro leads
Gemini 3 Pro: 57.6 (#2), GPT-5.5: 46.9 (#12)
| Benchmark | Gemini 3 Pro | GPT-5.5 |
|---|---|---|
| LMArena Vision | 1305 | 1297 |
| LMArena Document | 1434 | 1486 |
| GeoBench | 84% | — |
| VPCT | 91% | — |
| Blueprint-Bench 2 | — | 36.2% |
| Furniture Assembly | — | 44.2% |
Multilingual Too close to call
Gemini 3 Pro: 56.9 (#16), GPT-5.5: 56.4 (#20)
| Benchmark | Gemini 3 Pro | GPT-5.5 |
|---|---|---|
| LMArena Non-English | 1474 | 1467 |
| LMArena Chinese | 1523 | 1533 |
| LMArena French | 1492 | 1486 |
| LMArena German | 1515 | 1480 |
| LMArena Japanese | 1510 | 1498 |
| LMArena Korean | 1448 | 1460 |
| LMArena Russian | 1493 | 1473 |
| LMArena Spanish | 1470 | 1468 |
Instruction Following GPT-5.5 leads
Gemini 3 Pro: 76.3 (#45), GPT-5.5: 77.5 (#18)
| Benchmark | Gemini 3 Pro | GPT-5.5 |
|---|---|---|
| LMArena Instruction Following | 1458 | 1479 |
| IFEval | 87.7% | — |
Long Context GPT-5.5 leads
Gemini 3 Pro: 44.0 (#79), GPT-5.5: 48.3 (#12)
| Benchmark | Gemini 3 Pro | GPT-5.5 |
|---|---|---|
| LMArena Longer Query | 1471 | 1484 |
| CL-bench | 15.8% | — |
| CL-bench Life | — | 22.2% |
Writing & Preference GPT-5.5 leads
Gemini 3 Pro: 66.4 (#35), GPT-5.5: 72.7 (#13)
| Benchmark | Gemini 3 Pro | GPT-5.5 |
|---|---|---|
| LMArena Text | 1479 | 1472 |
| LMArena Creative Writing | 1482 | 1455 |
| EQ-Bench Creative Writing | 1525 | 1844 |
| LMArena Multi-Turn | 1484 | 1476 |
| WildBench | 85.9% | — |
| EQ-Bench 4 | — | 1315 |
Frequently asked questions
Is Gemini 3 Pro better than GPT-5.5?
GPT-5.5 is the stronger model overall, scoring 63.4 to 54.8 on the Noometry Index.
Is Gemini 3 Pro or GPT-5.5 better for coding?
GPT-5.5 scores higher on coding benchmarks: 58.2 versus 51.6 in the Noometry coding category.
How many benchmarks do Gemini 3 Pro and GPT-5.5 share?
47 benchmarks have published results for both models. Gemini 3 Pro has 67 scored results on Noometry and GPT-5.5 has 71.