OpenVLM HallusionBench - aAcc: leaderboard
Metric: Accuracy (%). Source: huggingface.co. 284 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5 | 77.2 |
| 2 | Gemini 2.5 Pro | 76 |
| 3 | GPT-5 Mini (2025-08-07) | 75.1 |
| 4 | GPT-5 Nano | 74 |
| 5 | GPT-4.5 | 73.4 |
| 6 | InternVL3-78B | 73.2 |
| 7 | GPT-4.1 (2025-04-14) | 73.1 |
| 8 | Qwen 2 VL 72B | 72.9 |
| 9 | InternVL2.5-78B | 72.9 |
| 10 | InternVL3-38B | 72.3 |
| 11 | Gemini 2.0 Flash | 72 |
| 12 | Gemini 1.5 Pro (002) | 71.9 |
| 13 | Claude 3.5 Sonnet (20241022) | 71.6 |
| 14 | GPT-4o ChatGPT | 71.4 |
| 15 | Gemini 1.5 Flash (002) | 69.5 |
Interactive version: theaggregate.ai/benchmark?slug=openvlm-hallusionbench-aacc · How It Works · Data refreshed daily, snapshot 2026-09-05.