ResearchQA (Open Paper) - Citation Accuracy: leaderboard

Metric: Citation accuracy (%; share of cited passages that are verifiable substrings of the paper, citation-grounded answers from the Open Paper chat-with-paper harness to a 100-row evenly spaced sample of 6,211 single-paper questions over 494 open-access papers, citations checked by a normalizing substring matcher against the extracted paper text; higher is better). Source: arxiv.org. Saturation forecast: Around December 2026. 8 models tracked.

Top models

#ModelScore
1Claude Haiku 4.584.5
2GLM-4.784
3Gemini 3 Flash (Preview)83.5
4GPT-5.483.4
5Gemini 3.1 Pro (Preview)81.7
6Claude Opus 4.780.9
7GPT-4.178.6
8GPT-OSS-120B70

Interactive version: theaggregate.ai/benchmark?slug=researchqa-open-paper-citation-accuracy · How It Works · Data refreshed daily, snapshot 2026-09-29.