LlamaParse
Industry-leading agentic document processing for complex documents.
LlamaIndex SDK Monthly Downloads
25M+
Total Funding
$27.5M
Free Tier
1k pages/day
About LlamaParse
LlamaParse provides a sophisticated, API-driven solution for parsing and extracting structured information from complex documents. It is specifically engineered to handle challenging layouts, embedded tables, and charts within PDFs, which traditional OCR methods often fail to interpret correctly. The service is a core component of the LlamaIndex ecosystem, designed to feed clean, structured data into Retrieval-Augmented Generation (RAG) pipelines. Developers can access LlamaParse via a REST API or dedicated Python and TypeScript SDKs, which handle asynchronous job polling. It operates on a credit-based system, offering a free tier for initial use and pay-as-you-go plans for scalable production workloads.
Pricing Model
Credit System
The platform operates on a credit-based system where actions like parsing or extraction consume credits.
Free Plan
Includes a generous free tier for developers, allowing up to 1,000 pages per day.
Pay-As-You-Go
Paid plans start with a Starter tier, allowing users to purchase credits for higher volume usage.
Cost Per Page
The number of credits per page varies based on the parsing mode and model used.
API & Developer Experience
Api Type
Asynchronous REST API. You submit a job and then poll for the result.
Sdks
Official SDKs for Python and TypeScript are available to simplify integration and handle job polling.
Authentication
Uses a LlamaCloud API key passed in the environment variables or client configuration.
Core Functionality
Provides agentic OCR, structured data extraction, classification, and indexing.