Model comparison
Gemini 3 Pro vs gpt-oss-120b
Gemini 3 Pro is the stronger model overall, scoring 54.7 to 36.0 on the Noometry Index.
Last verified . 38 shared benchmarks.
GoogleSummary
- They share 38 benchmarks with published results for both. Gemini 3 Pro scores higher in 8 categories and gpt-oss-120b in 1 category; 9 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Gemini 3 Pro leads 52.5 to 19.5.
- The biggest single-benchmark swing is SimpleBench: 76.4% for Gemini 3 Pro and 22.1% for gpt-oss-120b.
- gpt-oss-120b has downloadable open weights; the other is API-only.
Side by side
| Gemini 3 Pro | gpt-oss-120b | |
|---|---|---|
| Provider | OpenAI | |
| Noometry Index | 54.7 | 36.0 |
| Released | 2025-11-18 | 2025-08-05 |
| Weights | Proprietary | Open |
| Context window | — | 131K |
| Max output | — | 41K |
| Input $ / M tokens | — | $0.037 |
| Output $ / M tokens | — | $0.17 |
| Results tracked | 67 | 48 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Gemini 3 Pro leads
Gemini 3 Pro: 52.1 (#33), gpt-oss-120b: 33.2 (#251)
| Benchmark | Gemini 3 Pro | gpt-oss-120b |
|---|---|---|
| SWE-bench Verified (bash only) | 74.2% | 26% |
| WeirdML | 69.9% | 48.2% |
| LMArena Coding | 1482 | 1380 |
| ALE-Bench | 1,177 | 575.62 |
| AlgoTune | 1.83 | 1.41 |
| SWE-bench Verified | 72.9% | — |
| Aider Polyglot | — | 41.8% |
| LMArena WebDev | 1439 | — |
| SWE-bench Multilingual | 68.7% | — |
| SciCode | — | 36% |
| GSO | 18.6% | — |
Agentic & Tool Use Gemini 3 Pro leads
Gemini 3 Pro: 40.5 (#21), gpt-oss-120b: 12.1 (#145)
| Benchmark | Gemini 3 Pro | gpt-oss-120b |
|---|---|---|
| Terminal-Bench | 69.4% | 18.7% |
| METR Time Horizons | 71% | 56.6% |
| Vending-Bench 2 | 5,478 | -21.53 |
| APEX-Agents | — | 4.4% |
| Berkeley Function Calling Leaderboard | 72.5% | — |
| GDPval | 40.3% | — |
| Remote Labor Index | 1.3% | — |
| τ²-bench Airline | 80.5% | — |
| τ²-bench Banking | 18% | — |
| τ²-bench Retail | 75.9% | — |
| τ²-bench Telecom | 91% | — |
| DeepResearch Bench | 46.3% | — |
| BALROG | 58.1% | — |
| LMArena Search | 1207 | — |
Reasoning Gemini 3 Pro leads
Gemini 3 Pro: 52.5 (#28), gpt-oss-120b: 19.5 (#236)
| Benchmark | Gemini 3 Pro | gpt-oss-120b |
|---|---|---|
| SimpleBench | 76.4% | 22.1% |
| Kagi LLM Benchmark | 80.1% | 58.6% |
| CritPt | 6.9% | 1.1% |
| Chess Puzzles | 31% | 20% |
| LMArena Hard Prompts | 1481 | 1364 |
| Epoch Capabilities Index | 152.96 | 139.94 |
| ARC-AGI-2 | 31.1% | — |
| NYT Connections (extended) | 94.4% | — |
| ARC-AGI-1 | 75% | — |
| EnigmaEval | 18.2% | — |
| Mystery Game Puzzles | — | 2% |
| DTBench | — | 76.3% |
| LMCA | — | 22.1% |
| Surface Evolver Bench | — | 25% |
| ForecastBench | 61.2 | — |
Math gpt-oss-120b leads
Gemini 3 Pro: 48.8 (#50), gpt-oss-120b: 50.6 (#43)
| Benchmark | Gemini 3 Pro | gpt-oss-120b |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 91.4% | 88.9% |
| Omni-MATH | 55.5% | 68.8% |
| LMArena Math | 1475 | 1388 |
| MathArena Final-Answer Competitions | 67% | — |
| ProofBench | 20% | — |
| FrontierMath (Feb 2025 set) | 37.6% | — |
| FrontierMath Tier 4 (v1) | 18.8% | — |
Knowledge Gemini 3 Pro leads
Gemini 3 Pro: 64.3 (#15), gpt-oss-120b: 42.3 (#89)
| Benchmark | Gemini 3 Pro | gpt-oss-120b |
|---|---|---|
| GPQA Diamond | 92.6% | 75.8% |
| MMLU-Pro | 90.3% | 79.5% |
| Vectara Hallucination Rate | 13.6% | 14.2% |
| GPQA (HELM) | 80.3% | 68.4% |
| LMArena Expert | 1475 | 1356 |
| Humanity's Last Exam | 37.5% | — |
| Confabulations | — | 15.7% |
Multimodal Not comparable
Gemini 3 Pro: 56.9 (#1), gpt-oss-120b: —
| Benchmark | Gemini 3 Pro | gpt-oss-120b |
|---|---|---|
| LMArena Vision | 1305 | — |
| GeoBench | 84% | — |
| VPCT | 91% | — |
| LMArena Document | 1434 | — |
Multilingual Gemini 3 Pro leads
Gemini 3 Pro: 57.1 (#14), gpt-oss-120b: 48.2 (#142)
| Benchmark | Gemini 3 Pro | gpt-oss-120b |
|---|---|---|
| LMArena Non-English | 1474 | 1350 |
| LMArena Chinese | 1522 | 1384 |
| LMArena French | 1492 | 1370 |
| LMArena German | 1517 | 1355 |
| LMArena Japanese | 1513 | 1332 |
| LMArena Korean | 1449 | 1283 |
| LMArena Russian | 1493 | 1343 |
| LMArena Spanish | 1468 | 1388 |
Instruction Following Gemini 3 Pro leads
Gemini 3 Pro: 76.4 (#42), gpt-oss-120b: 69.5 (#168)
| Benchmark | Gemini 3 Pro | gpt-oss-120b |
|---|---|---|
| IFEval | 87.7% | 83.6% |
| LMArena Instruction Following | 1458 | 1318 |
Long Context Gemini 3 Pro leads
Gemini 3 Pro: 44.0 (#75), gpt-oss-120b: 31.4 (#273)
| Benchmark | Gemini 3 Pro | gpt-oss-120b |
|---|---|---|
| LMArena Longer Query | 1472 | 1319 |
| Fiction.LiveBench | — | 44.4% |
| CL-bench | 15.8% | — |
Writing & Preference Gemini 3 Pro leads
Gemini 3 Pro: 66.6 (#34), gpt-oss-120b: 46.7 (#213)
| Benchmark | Gemini 3 Pro | gpt-oss-120b |
|---|---|---|
| LMArena Text | 1479 | 1366 |
| LMArena Creative Writing | 1482 | 1274 |
| EQ-Bench Creative Writing | 1525 | 961 |
| WildBench | 85.9% | 84.5% |
| LMArena Multi-Turn | 1484 | 1340 |
| Short-Story Creative Writing | — | 77.1% |
Frequently asked questions
Is Gemini 3 Pro better than gpt-oss-120b?
Gemini 3 Pro is the stronger model overall, scoring 54.7 to 36.0 on the Noometry Index.
Is Gemini 3 Pro or gpt-oss-120b better for coding?
Gemini 3 Pro scores higher on coding benchmarks: 52.1 versus 33.2 in the Noometry coding category.
How many benchmarks do Gemini 3 Pro and gpt-oss-120b share?
38 benchmarks have published results for both models. Gemini 3 Pro has 67 scored results on Noometry and gpt-oss-120b has 48.