LlamaIndex is an open-source data framework designed for building applications that connect large language models with external data sources, with a particular focus on retrieval-augmented generation (RAG) and knowledge-grounded AI systems. Originally created by Jerry Liu in late 2022 under the name GPT Index, the framework has grown into one of the most widely used tools for building production RAG pipelines and LLM-powered data applications. LlamaIndex provides a comprehensive set of tools for ingesting data from over 160 sources including PDFs, databases, APIs, web pages, Slack, Notion, Google Drive, and many more through its LlamaHub connector ecosystem. The framework handles the complete RAG pipeline from data ingestion through indexing, retrieval, and response synthesis. Core components include document loaders and readers, node parsers for chunking and transforming documents, index structures for organizing data (vector, list, tree, keyword, and knowledge graph indices), retrievers for fetching relevant context, and response synthesizers for generating LLM responses grounded in retrieved data. LlamaIndex supports advanced retrieval strategies including hierarchical retrieval, recursive retrieval, fusion retrieval, auto-merging, and sentence window retrieval that go beyond simple vector similarity search to improve answer quality. The framework also provides agentic capabilities through LlamaIndex Workflows, enabling developers to build complex multi-step AI applications with tool use and reasoning. LlamaIndex integrates with all major LLM providers, embedding models, and vector stores. LlamaCloud is the companion managed service that provides managed ingestion and retrieval pipelines optimized for production use. The core framework is free and open-source under the MIT license, available in Python and TypeScript. LlamaCloud offers a free tier and paid plans starting at $399 per month for production workloads.
llamaindex.ai →