Evaluating Modern RAG: Textual, Multimodal, Dense, and Late Interaction Pipelines
2026-08-24 • Information Retrieval
Information Retrieval
AI summaryⓘ
The authors looked at different ways computers find and use information from documents. Usually, these methods only read text, which works well but struggles with things like tables and pictures. Newer methods use both images and text to improve results, but they require more computer power. The authors studied these trade-offs and created a way to help choose the best method based on how well it works and how much computing resources it needs.
Retrieval-augmented generationText-based retrievalMultimodal pipelinesVision-language modelsDense architecturesLate-interaction architecturesDocument understandingResource constraintsInformation retrieval
Authors
Emre Kuru, Mehmet Onur Keskin
Abstract
Retrieval-augmented generation (RAG) systems have traditionally relied on text-based pipelines that extract and retrieve information from documents. While efficient and lightweight, these approaches often struggle with documents where meaning is conveyed through layout, tables, and visual elements. Recent advances in multimodal pipelines, powered by vision-language models (VLMs), improve retrieval quality by jointly encoding visual and textual signals, but at increased computational and memory cost. We propose a quantitative, data-driven selection methodology that guides practitioners in choosing the most appropriate RAG pipeline for a given document corpus based on empirical effectiveness and resource constraints. We evaluate contemporary textual and multimodal pipelines, including dense and late-interaction architectures, analyze their trade-offs, and provide actionable guidance for balancing retrieval performance with system efficiency.