CiteVQA - Multi-Doc (1-Gold): 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 covers questions over several linked documents of which one holds the evidence, counting a question as correct only when a judge scores the answer at least 4 of 5 and the cited evidence is either judged relevant (at least 4 of 5) or recovers at least 60 percent of the key evidence; higher is better.

Metric: Strict attributed accuracy (%): share of questions with answer score at least 4 and either relevance at least 4 or evidence recall at least 0.6, on the multi-document questions with one gold document (25.7 percent of the set), 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

#ModelSettingScoreOverall rank
1Gemini 3.1 Pro (Preview)High79.7#102
2Gemini 3 Flash (Preview)High61.8#156
3GPT-5.4xHigh56.9#138
4Gemini 2.5 ProThinking, 32K48.9#280
5GPT-5.2xHigh38.8#190
6Seed 2.0 Pro33.5#226
7Gemma 4 31B29.8#390
8Qwen 3.5 27B22.6#380
9Qwen 3.5 397B A17B22.2#270
10Qwen 3 VL 235B A22B Instruct21.6#541
11Kimi K2.518.9#272
12Qwen 3.6 Plus18.5#245
13Qwen 3.5 9B17.7#838
14Qwen 3.5 122B A10B17#350
15Qwen 3 VL 32B Instruct16.2#607

No result here: #2 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=citevqa-multi-doc-1-gold · How It Works · Data refreshed daily, snapshot 2026-11-22.