CiteVQA - Answer Correctness: leaderboard
CiteVQA (2026) asks multimodal models to answer questions over long PDFs, 1,897 questions on 711 documents in seven domains, and to cite the exact page elements that support each answer as bounding boxes. This board is answer correctness alone, judged on a 0 to 5 scale and multiplied by 20, averaged over all questions; higher is better.
Metric: Answer correctness against the reference answer, judged on a 0-5 scale and multiplied by 20 (0-100), 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 February 2027. 20 models tracked. Updated 14 Nov 2026.
Top models
| # | Model | Setting | Score | Overall rank |
|---|---|---|---|---|
| 1 | GPT-5.4 | xHigh | 87.1 | #135 |
| 2 | Gemini 3.1 Pro (Preview) | High | 86.1 | #99 |
| 3 | Qwen 3.6 Plus | 85.9 | #238 | |
| 4 | Gemini 3 Flash (Preview) | High | 84.5 | #152 |
| 5 | Gemini 2.5 Pro | Thinking, 32K | 82.2 | #276 |
| 6 | Seed 2.0 Pro | 81.3 | #222 | |
| 7 | Qwen 3.5 397B A17B | 76.5 | #264 | |
| 8 | Qwen 3.5 35B A3B | 76.4 | #492 | |
| 9 | Qwen 3.5 27B | 75.6 | #375 | |
| 10 | Kimi K2.5 | 74.3 | #266 | |
| 11 | Qwen 3.5 122B A10B | 73.6 | #346 | |
| 12 | Qwen 3 VL 32B Instruct | 72.3 | #606 | |
| 13 | Qwen 3 VL 235B A22B Instruct | 72.3 | #532 | |
| 14 | GPT-5.2 | xHigh | 71.5 | #185 |
| 15 | Gemma 4 31B | 69.8 | #382 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=citevqa-answer-correctness · How It Works · Data refreshed daily, snapshot 2026-11-21.