CiteVQA - Evidence Recall: leaderboard
Metric: Recall (%) of the crucial evidence elements matched by a cited box at IoU 0.5, over all 1,897 questions, the model answers from the document pages and cites element-level bounding boxes, unified prompt, temperature 1.0, at most 4,096 output tokens, Qwen3-VL-235B-A22B judge; higher is better. Source: arxiv.org. Saturation forecast: Around January 2027. 20 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) (High) | 66 |
| 2 | Gemini 3 Flash (Preview) (High) | 45.4 |
| 3 | GPT-5.4 (xHigh) | 31 |
| 4 | Seed 2.0 Pro | 28.5 |
| 5 | GPT-5.2 (xHigh) | 18.2 |
| 6 | GLM-5V Turbo | 14.9 |
| 7 | Gemma 4 31B | 11.6 |
| 8 | Qwen 3 VL 235B A22B Instruct | 11.3 |
| 9 | Qwen 3.6 Plus | 7.7 |
| 10 | Qwen 3 VL 32B Instruct | 6.6 |
| 11 | Kimi K2.5 | 6.2 |
| 12 | Qwen 3.5 397B A17B | 5.4 |
| 13 | Qwen 3.5 27B | 5.3 |
| 14 | Qwen 3.5 122B A10B | 3.9 |
| 15 | Qwen 3 VL 30B A3B Instruct | 3.5 |
Interactive version: theaggregate.ai/benchmark?slug=citevqa-evidence-recall · How It Works · Data refreshed daily, snapshot 2026-10-07.