HELM VHELM - Image2Webpage: leaderboard
The "Image2Webpage" 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: EMD-Sim. Source: crfm.stanford.edu. Saturation forecast: Estimated already saturated. 18 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-4o (2024-05-13) | 66.48 | #326 |
| 2 | Claude 3 Sonnet (20240229) | 61.02 | #681 |
| 3 | Claude 3 Opus (20240229) | 37.85 | #413 |
| 4 | Claude 3 Haiku (20240307) | 0 | #777 |
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-image2webpage · How It Works · Data refreshed daily, snapshot 2026-10-23.