Gemma 4 31B (IT) (Reasoning): benchmark results
Provider: Google. Released 2026-04-02. Access: Open.
Aggregate Rating 1654 ± 14, rank #517 of 2201 rated models, from 148 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| SEA-HELM (Burmese) | 70.97 | Normalized Score | 100 |
| SEA-HELM (Burmese) - NLG | 57.95 | Normalized Score | 100 |
| SEA-HELM (Burmese) - NLR | 70.4 | Normalized Score | 100 |
| SEA-HELM (Burmese) - Safety | 78.31 | Normalized Score | 100 |
| SEA-HELM (Burmese) - Toxicity Detection | 69.99 | Normalized Score | 100 |
| SEA-HELM (English) - IFEval | 94.92 | Normalized Score | 100 |
| SEA-HELM (English) - LiveCodeBench v6 | 80.09 | Normalized Score | 100 |
| SEA-HELM (English) - MathArena | 56.21 | Normalized Score | 100 |
| SEA-HELM (Filipino) | 78.44 | Normalized Score | 100 |
| SEA-HELM (Filipino) - Belebele QA | 88.5 | Normalized Score | 100 |
| SEA-HELM (Filipino) - Causal Reasoning | 97.06 | Normalized Score | 100 |
| SEA-HELM (Filipino) - Cultural | 78.44 | Normalized Score | 100 |
Interactive version: theaggregate.ai/model?slug=gemma-4-31b-it-reasoning · How It Works · Data refreshed daily, snapshot 2026-11-12.