M2-Verify-Gen: leaderboard
Metric: Macro-F1 (%) of binary claim consistency verification (supported or refuted) on the M2-Verify-Gen test set (figures and captions from 16 arXiv categories, negation-perturbed claims), zero-shot; higher is better. Source: arxiv.org. Saturation forecast: Around January 2027. 8 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Qwen 2.5 VL 7B Instruct | 68.1 |
| 2 | InternVL3-8B | 63.8 |
| 3 | Mistral Small 3.1 | 63.1 |
| 4 | GPT-5 Mini | 59.9 |
| 5 | GPT-4o Mini | 56.6 |
| 6 | Phi-4 Multimodal Instruct | 45.7 |
| 7 | Pixtral-12B | 41.9 |
Interactive version: theaggregate.ai/benchmark?slug=m2-verify-gen · How It Works · Data refreshed daily, snapshot 2026-10-07.