Leaderboard / Model
Gemini 3.7 Flash
Google DeepMind · Closed weights · Released Aug 13, 2026 · $0.75 in · $3.75 out per 1M tokens
68Overall · rank 7
Score by job
Coding
72 · rank 10 · 3 of 5 benchmarks
| Benchmark | Result | Variant | z |
|---|---|---|---|
| DeepSWE | 65.5% | gemini-3.7-flash_medium | 0.53 |
| FrontierCode | 43.6% | gemini-3.7-flash_unknown | 0.42 |
| LMArena Coding | 1503 rating | gemini-3.7-flash-high | 0.58 |
Agents
63 · rank 6 · 2 of 4 benchmarks
| Benchmark | Result | Variant | z |
|---|---|---|---|
| APEX-Agents | 67.8% | gemini-3.7-flash_unknown | 1.47 |
| Remote Labor Index | 5.0% | gemini-3.7-flash_unknown | -0.54 |
Reasoning
83 · rank 4 · 2 of 4 benchmarks
| Benchmark | Result | Variant | z |
|---|---|---|---|
| GPQA Diamond | 94.8% | gemini-3.7-flash_high | 1.31 |
| ARC-AGI-2 | 84.6% | gemini-3.7-flash_high | 0.95 |
Math
60 · rank 12 · 5 of 5 benchmarks
| Benchmark | Result | Variant | z |
|---|---|---|---|
| FrontierMath Tiers 1–3 | 71.6% | gemini-3.7-flash_high | 0.25 |
| FrontierMath Tier 4 | 36.6% | gemini-3.7-flash_high | -0.18 |
| OTIS Mock AIME | 97.2% | gemini-3.7-flash_high | 0.41 |
| ProofBench | 58.0% | gemini-3.7-flash_unknown | 0.41 |
| LMArena Math | 1525 rating | gemini-3.7-flash-high | 1.68 |
Writing
82 · rank 7 · 2 of 2 benchmarks
| Benchmark | Result | Variant | z |
|---|---|---|---|
| LMArena Creative Writing | 1492 rating | gemini-3.7-flash-high | 1.22 |
| LMArena Instruction Following | 1484 rating | gemini-3.7-flash-high | 0.74 |
Design
50 · rank 17 · 1 of 1 benchmarks
| Benchmark | Result | Variant | z |
|---|---|---|---|
| LMArena WebDev | 1595 rating | gemini-3.7-flash-high | 0.36 |