Leaderboard / Model
Gemini 3 Pro
Google DeepMind · Closed weights · Released Nov 18, 2025 · Price not listed
33Overall · rank 27
Score by job
Coding
45 · rank 23 · 3 of 5 benchmarks
| Benchmark | Result | Variant | z |
|---|---|---|---|
| Terminal-Bench | 69.4% | gemini-3-pro-preview | -0.02 |
| SWE-bench Verified | 72.9% | gemini-3-pro-preview | -1.03 |
| LMArena Coding | 1482 rating | gemini-3-pro | -0.39 |
Agents
0 · rank 16 · 2 of 4 benchmarks
| Benchmark | Result | Variant | z |
|---|---|---|---|
| Remote Labor Index | 1.3% | gemini-3-pro-preview | -1.10 |
| METR Time Horizons | 3.7 h | gemini-3-pro-preview | -1.43 |
Reasoning
49 · rank 19 · 4 of 4 benchmarks
| Benchmark | Result | Variant | z |
|---|---|---|---|
| GPQA Diamond | 92.6% | gemini-3-pro-preview | 0.42 |
| ARC-AGI-2 | 31.1% | gemini-3-pro-preview | -1.51 |
| SimpleBench | 76.4% | gemini-3-pro-preview | 1.10 |
| Humanity’s Last Exam | 37.5% | gemini-3-pro-preview | -0.04 |
Math
22 · rank 33 · 3 of 5 benchmarks
| Benchmark | Result | Variant | z |
|---|---|---|---|
| OTIS Mock AIME | 91.4% | gemini-3-pro-preview | -1.02 |
| ProofBench | 20.0% | gemini-3-pro-preview | -1.09 |
| LMArena Math | 1476 rating | gemini-3-pro | -0.17 |
Writing
69 · rank 11 · 2 of 2 benchmarks
| Benchmark | Result | Variant | z |
|---|---|---|---|
| LMArena Creative Writing | 1481 rating | gemini-3-pro | 0.89 |
| LMArena Instruction Following | 1458 rating | gemini-3-pro | -0.23 |
Design
15 · rank 40 · 1 of 1 benchmarks
| Benchmark | Result | Variant | z |
|---|---|---|---|
| LMArena WebDev | 1439 rating | gemini-3-pro | -1.18 |