HELM VHELM - Pope: leaderboard
The "Pope" 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: PEM. Source: crfm.stanford.edu. Saturation forecast: Estimated already saturated. 48 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 2.0 Pro (Preview 02-05) | 89.6 | #305 |
| 2 | Gemini 2.0 Flash | 89.2 | #344 |
| 3 | Gemini 2.0 Flash (Preview) | 88.9 | #366 |
| 4 | Gemini 1.5 Flash (002) | 88.8 | #585 |
| 5 | Gemini 2.0 Flash Lite | 88.7 | #453 |
| 6 | Gemini 2.5 Pro (Preview 03-25) | 88.3 | #124 |
| 7 | Gemini 1.5 Pro (002) | 88 | #350 |
| 8 | GPT-4o (2024-05-13) | 87.9 | #326 |
| 9 | GPT-4.5 | 87 | #276 |
| 10 | GPT-4.1 Mini | 86.9 | #351 |
| 11 | O3 (2025-04-16) | 86.8 | #123 |
| 12 | GPT-4o (2024-08-06) | 86.6 | #335 |
| 13 | GPT-4o (2024-11-20) | 86.2 | #378 |
| 14 | GPT-4.1 (2025-04-14) | 85.9 | #250 |
| 15 | Llama 4 Maverick Instruct FP8 | 85.1 | #412 |
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-pope · How It Works · Data refreshed daily, snapshot 2026-10-23.