HELM VHELM - Safety: leaderboard
The "Safety" subtask of HELM VHELM. Stanford HELM's VHELM project, the holistic evaluation of vision-language models: image-grounded question answering, OCR, chart and diagram reading, spatial reasoning, robustness, bias, fairness and toxicity scenarios scored under one ha...
Metric: Mean win rate. Source: crfm.stanford.edu. Saturation forecast: Estimated already saturated. 48 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-4.1 Mini | 55.32 | #351 |
| 2 | Claude 3.5 Sonnet (20241022) | 55.32 | #291 |
| 3 | GPT-4.1 Nano | 55.32 | #735 |
| 4 | Gemini 2.0 Flash | 55.32 | #344 |
| 5 | Gemini 1.5 Flash (002) | 55.32 | #585 |
| 6 | Claude 3.5 Sonnet (20240620) | 55.32 | #298 |
| 7 | Llama 4 Scout Instruct | 55.32 | #542 |
| 8 | GPT-4o Mini (2024-07-18) | 55.32 | #561 |
| 9 | Gemini 2.5 Pro (Preview 03-25) | 55.32 | #124 |
| 10 | GPT-4.1 (2025-04-14) | 55.32 | #250 |
| 11 | GPT-4o (2024-08-06) | 55.32 | #335 |
| 12 | GPT-4o (2024-05-13) | 55.32 | #326 |
| 13 | O4 Mini (2025-04-16) | 55.32 | #181 |
| 14 | Llama 4 Maverick Instruct FP8 | 55.32 | #412 |
| 15 | GPT-4o (2024-11-20) | 55.32 | #378 |
No result here: #4 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=helm-vhelm-safety · How It Works · Data refreshed daily, snapshot 2026-10-23.