FlashRAG

Overview

FlashRAG is a Python toolkit for the reproduction and development of Retrieval Augmented Generation (RAG) research. FlashRAG is an efficient and modular open-source toolkit designed to assist researchers in reproducing and comparing existing RAG methods and developing their own algorithms within a unified framework.

Benchmarks and Algorithms

The toolkit includes:

• 36 pre-processed benchmark RAG datasets
• 23 state-of-the-art RAG algorithms, including 7 reasoning-based methods
• Support for reasoning-based methods that combine reasoning ability with retrieval (Search-o1, R1-Searcher, ReSearch)

Key Features

Multimodal RAG Support

Multimodal RAG support has been added, including MLLMs like Llava, Qwen, InternVL, and various multimodal retrievers with Clip architecture.

Reasoning Pipeline

A new paradigm that combines reasoning ability and retrieval, representing a significant advancement in RAG systems for complex reasoning tasks.

RAG Method Categories

RAG methods are categorized into four types based on their inference paths:

• Sequential: Sequential execution of RAG process
• Conditional: Implements different paths for different types of input queries
• Branching: Executes multiple paths in parallel, merging responses
• Loop: Iteratively performs retrieval and generation

Publication

The technical paper "FlashRAG: A Python Toolkit for Efficient RAG Research" was accepted to the Resource Track of the 2025 ACM Web Conference (WWW 2025).

Pricing

Free and open-source.