Gemma 4 31B IT — Benchmarks
Benchmark scores for Gemma 4 31B IT aggregated from public leaderboards, with how it ranks among open models. See hardware requirements for what you need to run it.
Overall rank: #46 of 78 open modelscomposite 41.9/100 across 9 benchmarks in 4 categories · methodology
Coding
| Benchmark | Score | Open rank | All models |
|---|---|---|---|
| SciCode | 43.4% | #21 / 62 | #97 / 161 |
| WeirdML | 52.3% | #9 / 46 | #65 / 161 |
| LMArena WebDev | 1364.2 | #28 / 48 | #87 / 123 |
| Surface Evolver Bench | 30.6% | #10 / 13 | #23 / 27 |
Instruction Following
| Benchmark | Score | Open rank | All models |
|---|---|---|---|
| LMArena Text | 1451.1 | #12 / 182 | #67 / 391 |
Knowledge
| Benchmark | Score | Open rank | All models |
|---|---|---|---|
| SimpleQA | 10.4% | #15 / 15 | #78 / 80 |
Math
| Benchmark | Score | Open rank | All models |
|---|---|---|---|
| AIME 2024/2025 | 73.3% | #24 / 82 | #117 / 270 |
Reasoning
| Benchmark | Score | Open rank | All models |
|---|---|---|---|
| Chess Puzzles | 5.0% | #29 / 65 | #152 / 205 |
| DTBench | 82.7% | #19 / 67 | #98 / 210 |
| LMCA | 39.3% | #9 / 53 | #86 / 172 |
| GPQA Diamond | 75.8% | #26 / 94 | #138 / 289 |
| CritPt | 1.4% | #27 / 62 | #106 / 172 |
Scores aggregated from public benchmark sources (each linked from the benchmark pages). llmrun does not run these benchmarks.