HELM VHELM - Hateful Memes: leaderboard
The "Hateful Memes" 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 | O4 Mini (2025-04-16) | 74.59 | #182 |
| 2 | GPT-4.5 | 73.91 | #276 |
| 3 | O3 (2025-04-16) | 73.09 | #124 |
| 4 | O1 (2024-12-17) | 70.77 | #153 |
| 5 | Gemini 2.5 Pro (Preview 03-25) | 69.26 | #125 |
| 6 | Llama 4 Scout Instruct | 68.58 | #541 |
| 7 | Llama 4 Maverick Instruct FP8 | 64.75 | #414 |
| 8 | Gemini 2.0 Flash | 64.21 | #345 |
| 9 | GPT-4.1 Mini | 63.11 | #358 |
| 10 | Claude 3.5 Sonnet (20241022) | 62.98 | #293 |
| 11 | Gemini 2.0 Pro (Preview 02-05) | 62.98 | #306 |
| 12 | Gemini 2.0 Flash Lite | 62.7 | #455 |
| 13 | Gemini 1.5 Pro (002) | 62.43 | #352 |
| 14 | GPT-4o (2024-11-20) | 61.48 | #379 |
| 15 | GPT-4 Turbo | 61.2 | #410 |
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-hateful-memes · How It Works · Data refreshed daily, snapshot 2026-10-24.