Open Source
Other RAGData FrameworkPythonAgentsVector Search
LlamaIndex
Python and TypeScript framework for building RAG pipelines and agents over structured and unstructured data. 160+ loaders, vector store integrations, and agentic workflows.
Install
$pip install llama-index About
LlamaIndex focuses on the data layer of LLM applications. Its loaders parse 160+ file types and API sources into a unified Document format, and its indexing layer stores them in any vector database with consistent query semantics.
The framework is particularly strong for production RAG — it includes query pipeline optimisation, re-ranking, metadata filters, and observability integrations out of the box.
Getting Started
pip install llama-index llama-index-llms-anthropic
from llama_index.core import VectorStoreIndex, SimpleDirectoryReader
from llama_index.llms.anthropic import Anthropic
docs = SimpleDirectoryReader("./data").load_data()
index = VectorStoreIndex.from_documents(docs)
query_engine = index.as_query_engine(llm=Anthropic(model="claude-sonnet-4-5"))
response = query_engine.query("Summarise the key findings")
print(response)Features
160+ data connectors (PDFs, Notion, Google Drive, databases, APIs)
Vector store integrations: Chroma, Qdrant, Weaviate, Pinecone, pgvector
Advanced RAG: hybrid search, re-ranking, metadata filtering, HyDE
LlamaCloud — managed data ingestion pipeline service
Structured data agents with SQL and pandas query engines
Multi-modal RAG with image understanding
Workflow API — event-driven agentic pipelines
TypeScript library (llama-index-ts) for JavaScript developers
100+ data connectors
Advanced RAG pipelines
Query engine abstraction
Structured data extraction
Agent & tool use
LlamaCloud managed service
References & Links
Project Info
Language Python
License MIT
Stars 40.0k ★
Forks 5.8k
Category Other
Topics
RAGData FrameworkPythonAgentsVector Search
Ready to try LlamaIndex?
Free and open source. Install in one command, works with Claude Code, Cursor, Windsurf, and more.
$pip install llama-index