Web Retrieval-Aware Chunking (W-RAC) for Efficient and Cost-Effective Retrieval-Augmented Generation Systems

TL;DR

This paper introduces W-RAC, a web-specific chunking framework for retrieval-augmented systems that reduces costs and improves scalability by decoupling text extraction from semantic chunking using LLMs.

Contribution

W-RAC is a novel, cost-efficient web document chunking method that separates content extraction from semantic grouping, enhancing scalability and reducing token costs.

Findings

W-RAC reduces chunking-related LLM costs by an order of magnitude.

W-RAC achieves comparable or better retrieval performance than traditional methods.

W-RAC improves system observability and eliminates hallucination risks.

Abstract

Retrieval-Augmented Generation (RAG) systems critically depend on effective document chunking strategies to balance retrieval quality, latency, and operational cost. Traditional chunking approaches, such as fixed-size, rule-based, or fully agentic chunking, often suffer from high token consumption, redundant text generation, limited scalability, and poor debuggability, especially for large-scale web content ingestion. In this paper, we propose Web Retrieval-Aware Chunking (W-RAC), a novel, cost-efficient chunking framework designed specifically for web-based documents. W-RAC decouples text extraction from semantic chunk planning by representing parsed web content as structured, ID-addressable units and leveraging large language models (LLMs) only for retrieval-aware grouping decisions rather than text generation. This significantly reduces token usage, eliminates hallucination risks, and improves system observability.Experimental analysis and architectural comparison demonstrate that W-RAC achieves comparable or better retrieval performance than traditional chunking approaches while reducing chunking-related LLM costs by an order of magnitude.