OpenVLM MMT-Bench - Deepfake Detection — leaderboard
Metric: Score (%). Source: huggingface.co. 207 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-4.1 Mini | 60 |
| 2 | Gemma 3 4B | 55 |
| 3 | GPT-4.1 (2025-04-14) | 55 |
| 4 | Llama 3.2 90B Vision Instruct | 55 |
| 5 | Qwen 2 VL 2B | 50 |
| 6 | GPT-4.1 Nano | 50 |
| 7 | Gemma 3 27B | 50 |
| 8 | Gemini 1.5 Flash | 50 |
| 9 | InternVL3-78B | 45 |
| 10 | InternVL2-8B | 45 |
| 11 | InternVL3-8B | 45 |
| 12 | Qwen 2 VL 72B | 45 |
| 13 | Qwen 2 VL 7B | 45 |
| 14 | InternVL3-38B | 45 |
| 15 | InternVL3-14B | 45 |
Interactive version: theaggregate.ai/benchmark?slug=openvlm-mmt-bench-deepfake-detection · How the rankings work · Data refreshed daily, snapshot 2026-07-22.