AI tools, scored for your job
Get paid to train AI
🤝 You're in the Free & Open Source library — community-built tools you can use without paying. A different concept from our scored AI directory.
Home › Free & Open Source › Logistics › py-idp
✦ Free & open sourceLogistics

py-idp

AI-enabled Python Intelligent Document Processing with 6-stage pipeline and Pydantic schemas.

AI Intelligent Document Processing for Python

Six-stage pipeline with Pydantic schemas, 12+ LLM backends, self-hosted OCR, and HITL review.

Free forever · open source · AGPL-3.0-or-later
✓ Light setup✓ Browser-based✓ Real-time collaboration
7.4
ZEKAI
Powerful Python IDP toolkit
Free & open source · Logistics
Ease of setup6
Usefulness9
Maintenance9.5
Community5
★ 297
GITHUB
Oct 2026
UPDATED
AGPL-3.0-or-later
LICENSE
🛠 Don't want to self-host? We can set it up →
⚡ Quick answer

py-idp is a Python toolkit for processing complex documents like invoices and contracts using LLMs. It suits backend engineers and supply chain developers building automated document extraction pipelines.

Runs on
Cross-platform
Hosting
Self-host
Languages
Python
License
AGPL-3.0-or-later

What it is

It provides a six-stage pipeline spanning parsing via Docling or pdfplumber, classification, extraction, confidence assessment, Pydantic validation, and Streamlit-based human-in-the-loop review. It supports auto-chunking for oversized documents, self-hosted Nanonets-OCR2-3B, and auto-schema discovery.

Who it's for

Suits Python developers automating logistics and supply chain document workflows. Does not suit teams lacking Python skills or those wanting a fully managed SaaS without self-hosting infrastructure.

Key features

Six-Stage Pipeline

Parse, classify, extract, assess, validate, and route documents programmatically.

Pydantic Schema Driven

Enforce strict typing and validation on all extracted document fields.

Multiple LLM Backends

Connects to 12+ international and China LLM providers plus self-hosted VLM options.

Human-in-the-Loop Review

Includes a Streamlit UI to review low-confidence extractions and save corrections.

How to start

Up and running in 3 steps

1
Run `pip install py-idp` to install the core package with mock support
2
Use `extract_one('invoice.pdf')` to run a zero-configuration test extraction
3
Configure your LLM backend and Pydantic schemas to scale up your pipeline
Real talk

Our honest verdict

What's great

  • Supports zero-config extraction out of the box with an in-tree mock backend
  • Built-in auto-chunking handles oversized documents with token-budget controls
  • Optional IBM Docling integration provides advanced PDF table extraction
  • Includes a built-in evaluation harness for measuring field F1 and schema-valid rates

Where it falls short

  • Requires writing Python code and configuring LLM API keys or local models
  • Running self-hosted OCR like Nanonets-OCR2-3B demands adequate GPU or Apple Silicon hardware
  • Production deployment requires setting up a FastAPI server and handling database state
Cost

Pricing & license

The core package is available under the AGPL-3.0-or-later license, and a commercial license is also available. The only cost is running your own infrastructure and LLM API usage.

Paid work

Open remote AI-training roles

880 open remote AI-training roles, updated November 15, 2026; most pay $20–$175 an hour. Zekai lists them free; the application goes to the employer’s platform.

See all open roles →
Compare

Alternatives

Get more

Use it, or have us set it up

📄 Free PDF: py-idp for Logistics
A plain-language guide to getting your first result. Sent to your inbox.
🔔 64 people want a full video tutorial — add your vote
🛠 Don't want to set it up yourself?
We'll install, host & configure py-idp for you — and train your team.
Get it done for you ↗
Questions

FAQ

Do I need an API key to start?

No, the core install ships with an in-tree mock backend that lets you test extractions without any API keys.

How does it handle multi-page documents?

It features built-in auto-chunking for oversized documents to fit within token budgets.

Can I review low-confidence extractions?

Yes, it includes a built-in Streamlit UI for human-in-the-loop review and correction persistence.

Radar

Latest AI news

✉️ The weekly free toolkit
The best new free & open-source tools for Logistics — one short email a week, in your language.
Free · 1 email/week · unsubscribe anytime
Get paid to train AI →

See Zekai first in Google