HELM VHELM - Math Vista - Daily Life (Free Form): leaderboard
VHELM (Stanford CRFM HELM for vision-language models, run v2.1.2) MathVista items at grade level daily life in free form format: mathematical reasoning over plots, charts, diagrams and photos; 221 questions, exact match. Percent correct, 0-100, higher is better.
Metric: Exact Match (%). Source: crfm.stanford.edu. Saturation forecast: Estimated already saturated. 48 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 2.5 Pro (Preview 03-25) | 72.85 | #126 |
| 2 | O4 Mini (2025-04-16) | 64.71 | #183 |
| 3 | O3 (2025-04-16) | 63.8 | #125 |
| 4 | Gemini 2.0 Pro (Preview 02-05) | 56.56 | #318 |
| 5 | Gemini 2.0 Flash | 52.94 | #351 |
| 6 | GPT-4.1 (2025-04-14) | 49.77 | #257 |
| 7 | O1 (2024-12-17) | 49.77 | #157 |
| 8 | Gemini 2.0 Flash (Preview) | 48.87 | #376 |
| 9 | GPT-4.5 | 48.42 | #278 |
| 10 | Claude 3.5 Sonnet (20241022) | 47.06 | #302 |
| 11 | Gemini 2.0 Flash Lite | 46.15 | #455 |
| 12 | GPT-4.1 Mini | 45.25 | #363 |
| 13 | Gemini 1.5 Pro (002) | 45.25 | #368 |
| 14 | Claude 3.5 Sonnet (20240620) | 42.08 | #308 |
| 15 | GPT-4o (2024-05-13) | 40.27 | #337 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=helm-vhelm-math-vista-daily-life-free-form · How It Works · Data refreshed daily, snapshot 2026-11-05.