Gemini 3.1 Pro (Preview) (Medium) — benchmark results
Provider: Google. Released 2026-02-19. Access: API.
Unified ELO 1964 ± 44, rank #24 of 1839 rated models, from 9 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Chess Bench LLM | 2048 | Lichess Rating | 99 |
| Context Arena MRCR (8-needle) | 48.5 | AUC@1M (%) | 97.7 |
| Context Arena MRCR (2-needle) | 87.7 | AUC@1M (%) | 95 |
| GIM | 1.85 | IRT ability (theta) | 91.1 |
| Context Arena MRCR (4-needle) | 62.8 | AUC@1M (%) | 90.2 |
| OckBench | 87.5 | Accuracy (%) | 86.4 |
| MathArena - ArXiv Math Jan 2026 | 67.39 | Accuracy (%) | 73.1 |
| MageBench Season 1 | 1602 | Rating (self-reported) | 58.8 |
| MageBench S1 | 1602 | Rating | 38.2 |
Interactive version: theaggregate.ai/model?slug=gemini-3-1-pro-preview-medium · How It Works · Data refreshed daily, snapshot 2026-08-05.