PleIAs RAG-Resources
PleIAs RAG-Resources gathers material aimed at RAG work into a single place on Hugging Face. Rather than a single benchmark, it is a collection, so it saves you the effort of tracking down suitable question-and-answer sets and corpora separately.
PleIAs is a French research lab focused on open datasets and models, so the collection reflects an emphasis on openly licensed, reusable data. Check the terms on each item, since the licence varies across the resources included.
Use it as a discovery tool: browse what is on offer, pull the pieces that match your domain, and combine them into an evaluation set for your own pipeline.
Related sources
FlashRAG Benchmark Datasets
A research toolkit that bundles standardised, pre-processed benchmark datasets for RAG evaluation across many scenarios. Every dataset comes in a single unified JSONL format, so you can swap benchmarks in and out without rewriting your data loading each time.
Open RAGBench (Vectara)
A RAG evaluation benchmark from Vectara built on 1,000 arXiv papers, with multimodal content extraction and 3,045 question-and-answer pairs across scientific domains. Designed to test retrieval and answer quality on real research documents rather than short, simplified passages.
RAG-Mini-Wikipedia
A small evaluation set that pairs 918 question-and-answer pairs with a matching corpus of 3,200 Wikipedia passages. Built specifically for testing RAG systems, it is small enough to run quick evaluation loops while still covering realistic retrieval questions.