LightRAG: Simple and Fast Retrieval-Augmented Generation

By Zirui Guo, Lianghao Xia, Yanhua Yu, Tu Ao, Chao Huang. Conference on Empirical Methods in Natural Language Processing, 2024

375 citations recorded by the source corpus.

Retrieval-Augmented Generation (RAG) systems enhance large language models (LLMs) by integrating external knowledge sources, enabling more accurate and contextually relevant responses tailored to user needs. However, existing RAG systems have significant limitations, including reliance on flat data representations and inadequate contextual awareness, which can lead to fragmented answers that fail to capture complex inter-dependencies. To address these challenges, we propose LightRAG, which incorporates graph structures into text indexing and retrieval processes. This innovative framework employs a dual-level retrieval system that enhances comprehensive information retrieval from both low-level and high-level knowledge discovery. Additionally, the integration of graph structures with vector representations facilitates efficient retrieval of related entities and their relationships, significantly improving response times while maintaining contextual relevance. This capability is further enhanced by an incremental update algorithm that ensures the timely integration of new data, allowing the system to remain effective and responsive in rapidly changing data environments. Extensive experimental validation demonstrates considerable improvements in retrieval accuracy and efficiency compared to existing approaches. We have made our LightRAG open-source and available at the link: https://github.com/HKUDS/LightRAG

Subfields: Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/1ea143c34b9bc359780f79ba4d68dee68bcc1129

Citations in the corpus

• ReMindRAG: Low-Cost LLM-Guided Knowledge Graph Traversal for Efficient RAG (2025, unknown): “• LightRAG [11].”
• A Matryoshka Hierarchical RAG for Efficient Multi-Hop Question Answering (2026, unknown): “LightRAG [6] combines this global view with a local, entity-based view through dual-level retrieval.”
• Constrained Dominant Sets for Multimodal Document Question Answering (2026, unknown): “…with consistency-constrained or iterative reasoning (VisDoMRAG [4], ViDoRAG [5]), entity-or knowledge-graph traversal (GraphRAG [29], LightRAG [7], MMGraphRAG [6], RAG-Anything [8]), multi-agent chain-of-thought (MA-RAG [9]), dual evolving graphs ( G 2 - Reader [10], previously the state of…”
• Min-Cost Flow Routing for Evidence Assembly in Long Multimodal Documents (2026, unknown): “GraphRAG (Edge et al., 2024) and LightRAG (Guo et al., 2024) construct entity–relation graphs, HippoRAG (Gutiérrez et al., 2024) applies personalised PageRank to a knowledge graph, and MMGraphRAG (Wan & Yu, 2025) incorporates multimodal nodes.”
• ASTRA-QA: A Benchmark for Abstract Question Answering over Documents (2026, unknown): “Topic Set-based Evaluation Head-to-head Evaluation Corpus QA Instances QA Refinement Answer 1 contains the [1,2,3] topics [4] and coversthe [a] To address these gaps, we introduce ASTRA-QA , a benchmark for AbSTRA ct QA over documents, as shown in Figure 1.”
• An Empirical Study of Knowledge Graph-Enhanced RAG for Information Security Compliance (2026, unknown): “Unlike traditional chunk-based systems, LightRAG constructs a semantic graph in which nodes represent clauses or concepts and edges capture relationships such as cross-references, thematic similarity, or hierarchical structure [30].”
• MG2-RAG: Multi-Granularity Graph for Multimodal Retrieval-Augmented Generation (2026, unknown): “– Text-Centric Graph Topology: Visual information is often converted into textual descriptions before constructing the graph [23, 65, 80].”
• From entity-centric to goal-oriented graphs: Enhancing LLM knowledge retrieval in minecraft (2025, unknown): “Therefore, GraphRAG and its variants were proposed to address this issue (Edge et al., 2025; Wu et al., 2024; Guo et al., 2024).”