Gemini 1.5 Pro (Preview 0514): benchmark results
Provider: Google. Access: API.
Aggregate Rating 1614 ± 22, rank #888 of 10916 rated models, from 140 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| HELM VHELM - Mmmu - Materials | 60 | Prefix Quasi-Exact Match (%) | 91.5 |
| HELM VHELM - Blink Perception - Object Localization | 91.8 | Prefix Quasi-Exact Match (%) | 89.4 |
| HELM VHELM - Blink Perception - Relative Reflectance | 67.91 | Prefix Quasi-Exact Match (%) | 87.2 |
| HELM VHELM - Exams V - Bulgarian (Natural Science) | 22.22 | Prefix Quasi-Exact Match (%) | 87.2 |
| HELM VHELM - Bias | 86.17 | Mean win rate | 86.2 |
| HELM VHELM - Pairs | 91.42 | PEM | 86.2 |
| HELM VHELM - A Okvqa Robustness | 87 | EM (Robustness) | 85.7 |
| HELM VHELM - Exams V - Hungarian (Natural Science) | 78.12 | Prefix Quasi-Exact Match (%) | 83 |
| HELM VHELM - Math Vista - High School (Free Form) | 87.1 | Exact Match (%) | 83 |
| HELM VHELM - Mm Star Perception - Coarse Perception | 83.2 | Prefix Quasi-Exact Match (%) | 81.9 |
| HELM VHELM - Mmmu - Physics | 75.86 | Prefix Quasi-Exact Match (%) | 81.9 |
| HELM VHELM - Mmmu - Math | 55.17 | Prefix Quasi-Exact Match (%) | 80.7 |
Interactive version: theaggregate.ai/model?slug=gemini-1-5-pro-preview-0514 · How It Works · Data refreshed daily, snapshot 2026-11-08.