Developers using LlamaIndex who need accurate parsing of complex PDFs, tables, and scanned documents.

LlamaParse is LlamaIndex's hosted document-parsing API that converts complex PDFs, spreadsheets, and images into structured Markdown or JSON for RAG pipelines.

Accessed via the LlamaCloud API using an API key from cloud.llamaindex.ai.

Introduced in 2024 with LlamaParse API v2 and new SDKs shipping in 2026.

It exists to solve high-fidelity parsing of tables, charts, and handwriting that plain OCR misses.

Integrates directly with LlamaIndex ingestion pipelines and is framework-agnostic for LangChain and others.

SaaS, proprietary, requires an API key; usage-based/credit billing; uploaded documents are processed on LlamaIndex's cloud.