Parse Examples
Runnable example tutorials demonstrating Parse for common document workflows — first parse, chart extraction, financial table extraction, custom prompts, batch processing, spreadsheets, and TypeScript/Node.js integration.
A collection of runnable tutorials for the most common Parse workflows. Each one walks through a real document end-to-end and shows the SDK calls in context. Most tutorials are Python notebooks (Jupyter, Colab); the TypeScript tutorial shows the same patterns from Node.js. If you’re new to Parse, the Getting Started page will give you the bare-bones first parse in 60 seconds.
Start here
Section titled “Start here” Quick Start: Parse a PDF & Interpret Outputs Python · Parse a real US Treasury financial report and interpret the four most common output views — text, markdown, items, and metadata. The best place to start if you're new to the Parse SDK.
Parse a PDF in TypeScript TypeScript · Parse a PDF from a Node.js script — install, set the API key, run the parse, walk the markdown output, and save it to disk. The TypeScript counterpart to the Python Quick Start.
By use case
Section titled “By use case”Structured extraction
Section titled “Structured extraction” Parse a Financial Report and Extract Every Table Walk every page of a real Treasury financial report, pull every table out into pandas DataFrames with source-page provenance, and turn on Cost Optimizer to keep costs down on long mixed-complexity documents.
Parse Embedded Charts and Analyze with Pandas Use specialized chart parsing to extract chart data from a financial report, then load the resulting tables into pandas for summaries, plots, and value counts.
Parse and Analyze Excel Spreadsheets Parse a multi-sheet XLSX template and (optionally) build a small RAG app over the rows with the LlamaIndex framework.
Prompt engineering
Section titled “Prompt engineering” Parse with Additional Prompts Steer Parse with a custom prompt to focus on the parts of a document you care about. Walks through extracting just the prices from a McDonald's receipt.
Bulk processing
Section titled “Bulk processing” Parse All PDFs in a Folder (Async) Batch-parse every PDF in a directory with controlled concurrency using asyncio and a semaphore. The pattern to use when you have hundreds or thousands of files.
More resources
Section titled “More resources”- Browse the SDK examples on GitHub — for more detailed Python SDK examples, see
run-llama/llama-cloud-py/tree/main/examples/parse. - Look up an option — see the Input Options, Output Options, and Processing Options groups for prose explanations of every knob.
- Pick the right tier — see the Tiers page for guidance on
fastvscost_effectivevsagenticvsagentic_plus. - Understand the configuration model — see the Configuration Model page for where every option lives in the request shape.