LlamaIndex logo

LlamaIndex

Free

provides a central interface to connect your LLM's with external data. ![GitHub Repo stars](https://img.shields.io/github/stars/jerryjliu/llama_index?style=social)

FreeFree tier
Type
Open Source
Company
LlamaIndex

About LlamaIndex

LlamaIndex is an open-source framework for building agentic applications powered by large language models. It provides a central interface to connect LLMs with external data, enabling advanced functionalities like document parsing, structured extraction, and retrieval-augmented generation (RAG). The framework offers over 300 integration packages (LlamaHub) for LLMs, embeddings, and vector stores, and supports flexible development through either a starter package (llama-index) or a customized core package (llama-index-core) with selected plugins. LlamaIndex also powers the enterprise platform LlamaParse, which includes agentic OCR, document parsing (130+ formats), Extract for structured data extraction, and Index for ingest and RAG pipelines.

Key Features

Open-source framework for building agentic applications
LlamaParse for agentic OCR and document parsing (130+ formats)
LlamaAgents for deployed document agents
Extract for structured data extraction from documents
Index for ingest, indexing, and RAG pipelines
Over 300 integration packages on LlamaHub for LLMs, embeddings, and vector stores
Flexible starter (llama-index) and customized (llama-index-core) packages
Namespaced architecture separating core and integration imports

Pros & Cons

Pros
  • Open-source with a large community (51k GitHub stars)
  • Extensive integration ecosystem with over 300 packages
  • Flexible architecture allows starter or customized setups
  • Includes enterprise-grade document parsing via LlamaParse
  • Well-documented with comprehensive guides and API references

Best For

Building agentic applications with LLMsDocument parsing and extraction across 130+ file formatsStructured data extraction from documentsRetrieval-augmented generation (RAG) pipelinesBuilding end-to-end document agents with workflowsIngesting and indexing data for LLM-powered search