Data ingestion
Document loaders, chunking, embeddings.
LlamaIndex
The data framework for LLM apps. Turn unstructured data into agent-queryable knowledge — document loaders, vector indices, query engines, tools, and multi-agent orchestration.
Choose your language
Section titled “Choose your language” 🐍 Python v0.14.21 · 100+ data connectors, ReAct agents, llama-agents package.
📘 TypeScript @llamaindex/workflow 1.1.25 · event-driven workflows for Node and browser.
What you’ll find in both guides
Section titled “What you’ll find in both guides”Vector indices
Summary, tree, vector store, list indices.
Query engines
RAG, subquestion, multi-step, router engines.
Agents
ReAct agents, function-calling agents, tool use.
Multi-agent
llama-agents (Python) / workflow orchestration (TypeScript).
Production
Observability, caching, scaling strategies.
Revision history
Section titled “Revision history”| Date | Version | Changes |
|---|---|---|
| 2026-05-04 | Py 0.14.21 / TS 1.1.25 | TS bumped 1.1.24 → 1.1.25; Python card corrected from 0.14.20 to 0.14.21 (stale reference). |
| 2026-04-21 | Py 0.14.21 / TS 1.1.24 | Versions corrected against installed packages; TS minimal example rewritten to real functional API (createWorkflow/workflowEvent). |
| 2026-04-21 | Py 0.14.20 / TS 1.1.4 | Parent index redesigned as language selector. |
| April 2026 | Py 0.14.20 / TS 1.1.4 | llama-agents stabilisation; scoped TS packages. |