Model comparison
Gemini 3 Pro vs o3
Gemini 3 Pro is the stronger model overall, scoring 54.8 to 47.5 on the Noometry Index.
Last verified . 51 shared benchmarks.
GoogleSummary
- They share 51 benchmarks with published results for both. Gemini 3 Pro scores higher in 8 categories and o3 in 2 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Gemini 3 Pro leads 52.5 to 32.0.
- The biggest single-benchmark swing is VPCT: 91% for Gemini 3 Pro and 52% for o3.
Side by side
| Gemini 3 Pro | o3 | |
|---|---|---|
| Provider | OpenAI | |
| Noometry Index | 54.8 | 47.5 |
| Released | 2025-11-18 | 2025-04-16 |
| Weights | Proprietary | Proprietary |
| Context window | — | 200K |
| Max output | — | 100K |
| Input $ / M tokens | — | $2 |
| Output $ / M tokens | — | $8 |
| Results tracked | 67 | 63 |
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), o3: 46.8 (#64)
| Benchmark | Gemini 3 Pro | o3 |
|---|---|---|
| SWE-bench Verified | 72.9% | 62.3% |
| SWE-bench Verified (bash only) | 74.2% | 58.4% |
| GSO | 18.6% | 8.8% |
| WeirdML | 69.9% | 52.4% |
| LMArena Coding | 1481 | 1408 |
| ALE-Bench | 1,177 | 933.55 |
| Aider Polyglot | — | 81.3% |
| LMArena WebDev | 1440 | — |
| SWE-bench Multilingual | 68.7% | — |
| CadEval | — | 74% |
| AlgoTune | 1.83 | — |
Agentic & Tool Use Gemini 3 Pro leads
Gemini 3 Pro: 40.6 (#23), o3: 34.5 (#44)
| Benchmark | Gemini 3 Pro | o3 |
|---|---|---|
| Berkeley Function Calling Leaderboard | 72.5% | 63% |
| GDPval | 40.3% | 30.8% |
| DeepResearch Bench | 46.3% | 45.2% |
| LMArena Search | 1207 | 1144 |
| METR Time Horizons | 71% | 65.4% |
| Terminal-Bench | 69.4% | — |
| Remote Labor Index | 1.3% | — |
| τ²-bench Airline | 80.5% | — |
| τ²-bench Banking | 18% | — |
| τ²-bench Retail | 75.9% | — |
| τ²-bench Telecom | 91% | — |
| OSWorld | — | 23% |
| BALROG | 58.1% | — |
| Vending-Bench 2 | 5,478 | — |
Reasoning Gemini 3 Pro leads
Gemini 3 Pro: 52.5 (#31), o3: 32.0 (#78)
| Benchmark | Gemini 3 Pro | o3 |
|---|---|---|
| ARC-AGI-2 | 31.1% | 6.5% |
| SimpleBench | 76.4% | 53.1% |
| Kagi LLM Benchmark | 80.1% | 67.6% |
| ARC-AGI-1 | 75% | 60.8% |
| CritPt | 6.9% | 1.4% |
| Chess Puzzles | 31% | 38% |
| EnigmaEval | 18.2% | 13.1% |
| LMArena Hard Prompts | 1480 | 1402 |
| Epoch Capabilities Index | 152.92 | 146.86 |
| ForecastBench | 61.2 | 62.5 |
| NYT Connections (extended) | 94.4% | — |
| Mystery Game Puzzles | — | 29% |
| DTBench | — | 84.8% |
| LMCA | — | 39.7% |
Math Too close to call
Gemini 3 Pro: 49.9 (#59), o3: 50.2 (#58)
| Benchmark | Gemini 3 Pro | o3 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 91.4% | 84.4% |
| Omni-MATH | 55.5% | 71.4% |
| LMArena Math | 1476 | 1426 |
| FrontierMath (Feb 2025 set) | 37.6% | 18.7% |
| FrontierMath Tier 4 (v1) | 18.8% | 2.1% |
| FrontierMath (Tiers 1-3) | — | 33.3% |
| MathArena Final-Answer Competitions | 67% | — |
| ProofBench | 20% | — |
| MATH Level 5 | — | 97.8% |
Knowledge Gemini 3 Pro leads
Gemini 3 Pro: 64.4 (#16), o3: 54.6 (#52)
| Benchmark | Gemini 3 Pro | o3 |
|---|---|---|
| GPQA Diamond | 92.6% | 81.8% |
| Humanity's Last Exam | 37.5% | 20.3% |
| MMLU-Pro | 90.3% | 85.9% |
| GPQA (HELM) | 80.3% | 75.3% |
| LMArena Expert | 1475 | 1402 |
| SimpleQA Verified | — | 49.4% |
| Confabulations | — | 14.4% |
| Vectara Hallucination Rate | 13.6% | — |
Multimodal Gemini 3 Pro leads
Gemini 3 Pro: 57.6 (#2), o3: 41.4 (#36)
| Benchmark | Gemini 3 Pro | o3 |
|---|---|---|
| LMArena Vision | 1305 | 1214 |
| GeoBench | 84% | 74% |
| VPCT | 91% | 52% |
| LMArena Document | 1434 | — |
Multilingual Gemini 3 Pro leads
Gemini 3 Pro: 56.9 (#16), o3: 51.7 (#105)
| Benchmark | Gemini 3 Pro | o3 |
|---|---|---|
| LMArena Non-English | 1474 | 1401 |
| LMArena Chinese | 1523 | 1437 |
| LMArena French | 1492 | 1430 |
| LMArena German | 1515 | 1420 |
| LMArena Japanese | 1510 | 1403 |
| LMArena Korean | 1448 | 1370 |
| LMArena Russian | 1493 | 1406 |
| LMArena Spanish | 1470 | 1395 |
Instruction Following Gemini 3 Pro leads
Gemini 3 Pro: 76.3 (#45), o3: 72.8 (#127)
| Benchmark | Gemini 3 Pro | o3 |
|---|---|---|
| IFEval | 87.7% | 86.9% |
| LMArena Instruction Following | 1458 | 1368 |
Long Context o3 leads
Gemini 3 Pro: 44.0 (#79), o3: 53.3 (#6)
| Benchmark | Gemini 3 Pro | o3 |
|---|---|---|
| CL-bench | 15.8% | 17.8% |
| LMArena Longer Query | 1471 | 1372 |
| Fiction.LiveBench | — | 88.9% |
Writing & Preference Gemini 3 Pro leads
Gemini 3 Pro: 66.4 (#35), o3: 63.5 (#64)
| Benchmark | Gemini 3 Pro | o3 |
|---|---|---|
| LMArena Text | 1479 | 1410 |
| LMArena Creative Writing | 1482 | 1359 |
| EQ-Bench Creative Writing | 1525 | 1676 |
| WildBench | 85.9% | 86.1% |
| LMArena Multi-Turn | 1484 | 1405 |
| Short-Story Creative Writing | — | 83.9% |
Frequently asked questions
Is Gemini 3 Pro better than o3?
Gemini 3 Pro is the stronger model overall, scoring 54.8 to 47.5 on the Noometry Index.
Is Gemini 3 Pro or o3 better for coding?
Gemini 3 Pro scores higher on coding benchmarks: 51.6 versus 46.8 in the Noometry coding category.
How many benchmarks do Gemini 3 Pro and o3 share?
51 benchmarks have published results for both models. Gemini 3 Pro has 67 scored results on Noometry and o3 has 63.