Leaderboard / Model
Gemini 3 Flash
Google DeepMind · Closed weights · Released Dec 17, 2025 · Price not listed
30Overall · rank 29
Score by job
Coding
38 · rank 26 · 3 of 5 benchmarks
| Benchmark | Result | Variant | z |
|---|---|---|---|
| Terminal-Bench | 64.3% | gemini-3-flash-preview | -0.46 |
| SWE-bench Verified | 75.4% | gemini-3-flash-preview | -0.28 |
| LMArena Coding | 1462 rating | gemini-3-flash | -1.39 |
Agents
Not enough data · 0 of 4 benchmarks
Reasoning
22 · rank 33 · 3 of 4 benchmarks
| Benchmark | Result | Variant | z |
|---|---|---|---|
| GPQA Diamond | 89.4% | gemini-3-flash-preview_high | -0.87 |
| ARC-AGI-2 | 33.6% | gemini-3-flash-preview | -1.40 |
| SimpleBench | 61.1% | gemini-3-flash-preview | -0.43 |
Math
22 · rank 33 · 5 of 5 benchmarks
| Benchmark | Result | Variant | z |
|---|---|---|---|
| FrontierMath Tiers 1–3 | 51.2% | gemini-3-flash-preview | -1.16 |
| FrontierMath Tier 4 | 17.1% | gemini-3-flash-preview | -1.00 |
| OTIS Mock AIME | 95.6% | gemini-3-flash-preview_high | -0.00 |
| ProofBench | 15.0% | gemini-3-flash-preview | -1.29 |
| LMArena Math | 1474 rating | gemini-3-flash | -0.24 |
Writing
55 · rank 29 · 2 of 2 benchmarks
| Benchmark | Result | Variant | z |
|---|---|---|---|
| LMArena Creative Writing | 1458 rating | gemini-3-flash | 0.22 |
| LMArena Instruction Following | 1438 rating | gemini-3-flash | -1.00 |
Design
15 · rank 40 · 1 of 1 benchmarks
| Benchmark | Result | Variant | z |
|---|---|---|---|
| LMArena WebDev | 1438 rating | gemini-3-flash | -1.19 |