HELM VHELM - Crossmodal 3600 - Chinese Locale: leaderboard
Metric: Prometheus Vision Rating (1-5). Source: crfm.stanford.edu. 48 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | O4 Mini (2025-04-16) | 4.05 |
| 2 | GPT-4.1 Mini | 3.96 |
| 3 | Claude 3.5 Sonnet (20241022) | 3.82 |
| 4 | GPT-4.1 (2025-04-14) | 3.78 |
| 5 | GPT-4 Turbo | 3.74 |
| 6 | O3 (2025-04-16) | 3.68 |
| 7 | Gemini 2.5 Pro (Preview 03-25) | 3.65 |
| 8 | GPT-4.1 Nano | 3.61 |
| 9 | Claude 3.5 Sonnet (20240620) | 3.55 |
| 10 | GPT-4o Mini (2024-07-18) | 3.54 |
| 11 | GPT-4o (2024-05-13) | 3.54 |
| 12 | GPT-4o (2024-11-20) | 3.52 |
| 13 | Claude 3 Opus (20240229) | 3.48 |
| 14 | O1 (2024-12-17) | 3.48 |
| 15 | Gemini 1.5 Pro (002) | 3.47 |
Interactive version: theaggregate.ai/benchmark?slug=helm-vhelm-crossmodal-3600-chinese-locale · How It Works · Data refreshed daily, snapshot 2026-09-19.