HELM VHELM - Exams V - Italian (Social Sciences): leaderboard
Metric: Prefix Quasi-Exact Match (%). Source: crfm.stanford.edu. 48 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | O3 (2025-04-16) | 100 |
| 2 | Gemini 2.5 Pro (Preview 03-25) | 96.77 |
| 3 | O1 (2024-12-17) | 93.55 |
| 4 | Gemini 2.0 Flash | 87.1 |
| 5 | O4 Mini (2025-04-16) | 87.1 |
| 6 | GPT-4o (2024-05-13) | 83.87 |
| 7 | GPT-4.1 Mini | 80.65 |
| 8 | GPT-4.1 (2025-04-14) | 80.65 |
| 9 | Gemini 2.0 Flash (Preview) | 80.65 |
| 10 | GPT-4.5 | 80.65 |
| 11 | GPT-4o (2024-08-06) | 77.42 |
| 12 | GPT-4o (2024-11-20) | 77.42 |
| 13 | Gemini 2.0 Pro (Preview 02-05) | 77.42 |
| 14 | Gemini 2.0 Flash Lite | 77.42 |
| 15 | Claude 3.5 Sonnet (20240620) | 74.19 |
Interactive version: theaggregate.ai/benchmark?slug=helm-vhelm-exams-v-italian-social-sciences · How It Works · Data refreshed daily, snapshot 2026-09-19.