Gemini 3.1 Pro (Preview) (High): benchmark results
Gemini 3.1 Pro (Preview) evaluated at the high reasoning-effort setting. Provider: Google. Released 2026-02-19. Access: API.
Unified ELO 1731 ± 1, rank #51 of 1761 rated models, from 117 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AudioMC | 66.8 | Score (self-reported) | 100 |
| Chess Bench LLM | 2081 | Lichess Rating | 100 |
| LiveBench Summarize | 88.23 | Score | 100 |
| MMAU | 82.5 | Avg Test (self-reported) | 100 |
| SEAL - Humanity's Last Exam (Text Only) | 47.31 | Score | 100 |
| SuperCLUE-AdsVU - Creative Element Recognition | 97.78 | Score | 100 |
| Vals AI AIME | 98.12 | Accuracy (%) | 100 |
| Vals AI GPQA | 95.45 | Accuracy (%) | 100 |
| NonoBench | 63.3 | Overall Accuracy (%) | 98.9 |
| UGI - Natural Intelligence | 74.94 | NatInt Score | 98.9 |
| Vals AI MedCode | 59.06 | Accuracy (%) | 98.9 |
| SimpleQA Verified | 73.5 | Accuracy (%) | 98.7 |
Interactive version: theaggregate.ai/model?slug=gemini-3-1-pro-preview-high · How It Works · Data refreshed daily, snapshot 2026-09-05.