CLI Quickstart

The ade CLI parses documents into Markdown and extracts fields you define in a schema, recording where on the page each value came from. It accepts PDFs and images, saves every result on your machine, and consumes no credits when you re-run an identical command.

Install the CLI

The installer downloads a self-contained binary for your platform and puts it in ~/.ade/bin. No Python or other runtime is needed.

curl -fsSL https://raw.githubusercontent.com/landing-ai/ade-cli/main/scripts/install.sh | sh

irm https://raw.githubusercontent.com/landing-ai/ade-cli/main/scripts/install.ps1 | iex

curl -fsSL https://raw.githubusercontent.com/landing-ai/ade-cli/main/scripts/install.cmd -o install.cmd && install.cmd && del install.cmd

Log In

Log in with your ADE account. Your browser opens to authenticate, and the CLI stores the credential. For more login options, see Authentication.

ade login

Parse a Document

Download the sample patient billing statement, then point ade parse at it. The run creates a folder and saves the parsed Markdown output (parse.md) and related files in it. To run more commands on this document, like extract or view, use the job item ID.

Command

ade parse -d patient-statement-brackenmill.pdf

Output

parse completed · 11s
Parsed /Users/you/documents/patient-statement-brackenmill.pdf -> job item bee989875c369fc5
  run:     parse-01kza8bh5jtazk6tdb2teacank
  model:   dpt-3-pro-20260710
  pages:   2 (0 failed)
  credits: 4.5 (priority)
  saved:   ~/.ade/jobs/bee989875c369fc5/  (parse.json, parse.md, elements.json)
  next:    ade view bee98987 --open   ·   ade extract bee98987 --schema <schema.json>

Extract the Fields You Need

A schema tells the extraction which fields to pull and how to recognize them. Download the sample schema, or save the block below as schema-patient-statement.json. To create your own later, build it visually in the Playground and download the JSON, or write the schema with Extraction Schema (JSON).

schema-patient-statement.json

{
  "description": "Root object",
  "type": "object",
  "properties": {
    "patient_name": {
      "description": "The full name of the patient receiving medical services.",
      "x-alternativeNames": [
        "Patient Name",
        "Name of Patient",
        "Recipient Name"
      ],
      "type": "string"
    },
    "medical_record_number": {
      "description": "A unique identifier assigned to the patient's medical records.",
      "x-alternativeNames": [
        "Medical Record Number",
        "Patient ID",
        "Chart Number"
      ],
      "type": "string"
    },
    "balance_due": {
      "description": "The total outstanding amount that the patient or guarantor needs to pay.",
      "x-alternativeNames": [
        "Balance Due",
        "Amount Due",
        "Pay This Amount"
      ],
      "type": "number"
    }
  },
  "required": [
    "patient_name",
    "medical_record_number",
    "balance_due"
  ]
}

Run the Extraction

Run the extraction against your parse, using the job item ID from the parse output. The run saves the extracted values (extract.json) and a record of where each value was found (evidence.json) in a folder of its own.

Command

ade extract bee989875c369fc5 --schema schema-patient-statement.json

Output

extract completed · 16s
Extracted /Users/you/documents/patient-statement-brackenmill.pdf -> job item 872cb75827d3f144
  run:      extract-01kza8qwyjtd1z4dbs6y1p31dz
  model:    extract-20260710
  fields:   3 (all grounded)
  evidence: grounded (field->box join from stored artifacts)
  credits:  2.2 (priority)
  saved:    ~/.ade/jobs/872cb75827d3f144/  (extract.json, evidence.json)
  next:     ade history list --json   ·   ade view 872cb758 --open

See Where Every Value Came From

Open the extraction in your browser. The viewer shows the extracted values beside the original pages, and selecting a field highlights the exact spot on the document it came from:

Command

ade view 872cb75827d3f144

Output

Viewer for job item 872cb75827d3f144 (extract) -> ~/.ade/jobs/872cb75827d3f144/view.html (built)
  sidebar: building missing sibling viewers in the background
  opened in browser: file:///Users/you/.ade/jobs/872cb75827d3f144/view.html

The ade viewer in a browser, showing the parsed patient statement with its elements outlined on the left, and the extraction results, per-field metadata, and schema fields on the right.

Common Tasks

Each command below takes a job item ID, or any unambiguous prefix of one. AI agents: ade help --json returns this whole surface as one JSON object.

To do this	Run	Learn more
Parse a local file	`ade parse -d document.pdf`	Parse Documents
Parse at half the credits	`ade parse -d document.pdf --tier standard`	Parse Documents
Parse only some pages	`ade parse -d document.pdf --pages 1,3-5`	Parse Documents
Extract fields from a parse	`ade extract JOB_ID --schema schema.json`	Extract Fields
Get the result as JSON for a script	`ade extract JOB_ID --schema schema.json --json`	Scripting and Automation
See your past runs	`ade history list`	CLI Concepts
Search a parsed document	`ade find JOB_ID "term"`	Find, Crop, and View
Open a document in the viewer	`ade view JOB_ID`	Find, Crop, and View
Save an element as an image	`ade crop JOB_ID --element-id ID`	Find, Crop, and View
Capture an ID for a script	`ade parse -d document.pdf --id-only`	Scripting and Automation
Check which account you are using	`ade auth status`	Authentication

Get Started

Use the CLI

Reference

Install the CLI

Log In

Parse a Document

Extract the Fields You Need

Run the Extraction

See Where Every Value Came From

Common Tasks

​Install the CLI

​Log In

​Parse a Document

​Extract the Fields You Need

​Run the Extraction

​See Where Every Value Came From

​Common Tasks

Install the CLI

Log In

Parse a Document

Extract the Fields You Need

Run the Extraction

See Where Every Value Came From

Common Tasks