landingai-ade

What it is and what it does

landingai-ade is the official Python client for LandingAI's Agentic Document Extraction API. It parses PDFs and images into structured Markdown with pixel-level grounding (bounding boxes), then extracts typed fields by applying a Pydantic model or JSON schema to that Markdown. The library provides both synchronous and asynchronous clients with identical interfaces, automatic retry logic with exponential backoff, and a job queue for processing large documents asynchronously. Responses are fully typed and include extraction metadata showing where each field's value came from in the source document.

The package is designed for workflows where you need to reliably extract structured data from unstructured documents—invoices, forms, contracts, receipts—without building your own OCR or layout analysis. It handles partial failures gracefully (HTTP 206 responses when some pages fail), supports optional disk persistence via save_to, and requires only an API key to authenticate. The v2 API (powered by DPT-3 models) is the current recommended interface; v1 methods remain supported for backward compatibility.

Use it for:

• Extract invoice line items, totals, and dates from PDF receipts and invoices into a typed Pydantic model.
• Parse form submissions or scanned documents into structured JSON for database ingestion or downstream processing.
• Build a document triage system that parses PDFs asynchronously and extracts metadata (sender, date, document type) at scale.
• Validate extracted fields against a schema and retrieve the pixel coordinates of each field for audit or UI highlighting.
• Process large multi-page documents via the job API without blocking on synchronous request timeouts.

Worth the install?

AI-flagged interpretation of the facts on this page — verify before relying

Converts PDFs and images into structured Markdown and extracts typed fields using JSON Schema or Pydantic models via the LandingAI Agentic Document Extraction API.

Yes. The package is actively maintained, has low install friction, carries a permissive license, and solves a concrete problem—structured extraction from unstructured documents—without requiring you to build OCR or layout parsing yourself. Suitable for production use if you have a LandingAI API account and your document extraction workflow aligns with the supported models. No known security vulnerabilities.

Install

landingai-ade on PyPI

pip install landingai-ade

uv add landingai-ade

poetry add landingai-ade

Installing landingai-ade

Before you install

Low friction: pure Python wheel with six lightweight runtime dependencies (anyio, distro, httpx, pydantic, sniffio, typing-extensions). Active maintenance—released 10 days ago with 1026 GitHub stars and commits through 2026-08-11.

License in practice

Apache-2.0 permissive license allows commercial and private use with minimal restrictions; suitable for most production and proprietary projects.

Quickstart

pip install landingai-ade

from pathlib import Path
from pydantic import BaseModel, Field
from landingai_ade import LandingAIADE

class Invoice(BaseModel):
    invoice_number: str = Field(description="The invoice number")
    total: str = Field(description="Invoice grand total")

client = LandingAIADE()  # reads VISION_AGENT_API_KEY
parsed = client.v2.parse(document=Path("invoice.pdf"))
result = client.v2.extract(schema=Invoice, markdown=parsed.markdown)
print(result.extraction)

Requires VISION_AGENT_API_KEY environment variable or explicit apikey parameter; LandingAI API account needed.

Verify before relying

• Exact rate limits, quota, or cost structure for the underlying LandingAI API service.
• Whether the DPT-3 model family supports all document types (e.g., handwritten, non-Latin scripts, scanned images).
• Latency and throughput characteristics for typical document sizes and extraction schemas.

Package facts

Evidence: landingai_ade-1.16.0-py3-none-any.whl

Tags

More Python Modules packages