langchain-pdfluent

A LangChain document loader for PDFluent. It reads a PDF with PDFluent's structured extraction and hands LangChain Documents in reading order, with the page and box each came from in its metadata so an answer can cite its source. Everything runs locally. Your documents stay on your machine.

Install

pip install langchain-pdfluent

It needs Python 3.9 or later and pdfluent 1.2.0 or later, which pip pulls in automatically. It also depends on langchain-core (MIT).

Quick start

from langchain_pdfluent import PDFluentLoader

docs = PDFluentLoader("report.pdf").load()
print(docs[0].page_content, docs[0].metadata["page"])

A runnable notebook, load and split, is in examples/load_and_split.ipynb.

Modes

mode decides what one Document is.

Blocks mode

One document per paragraph, heading, list item or table (a table is one document), in reading order. This is the default.

from langchain_pdfluent import PDFluentLoader

loader = PDFluentLoader("report.pdf")
for doc in loader.lazy_load():
    print(doc.metadata["kind"], doc.metadata["page"], doc.metadata["bbox"])

Markdown mode

One document per page, as markdown. Pass quality="balanced" for the layout-model tier (see below).

from langchain_pdfluent import PDFluentLoader

pages = PDFluentLoader("report.pdf", mode="markdown").load()
print(len(pages), "pages")
print(pages[0].page_content[:200])

Document mode

One document for the whole selection: the markdown of every selected page, joined by a blank line. metadata["pages"] lists the pages it covers.

from langchain_pdfluent import PDFluentLoader

whole = PDFluentLoader("report.pdf", mode="document", pages=[0, 2]).load()
print(whole[0].metadata["pages"])

Quality and pages

quality is "fast" (default) or "balanced", the layout-model tier of pdfluent 1.2. The balanced tier assembles markdown, so it goes with mode="markdown" or mode="document"; with mode="blocks" the engine refuses it and says so. It needs the model weights, installed once with the pdfluent command-line tool:

pdfluent models install layout tables

from langchain_pdfluent import PDFluentLoader

PDFluentLoader("report.pdf", mode="markdown", quality="balanced").load()

pages=[0, 3] loads exactly those pages (0-based); first_page and page_count load a window.

Options

Metadata

bbox, toc_items and tables are lists. A vector store that only accepts scalar metadata, such as Chroma, needs filter_complex_metadata from langchain_community over the documents first.

Benchmarks

Extraction quality is measured in the public pdfluent-bench repository (link follows when it is public).

Licence

AGPL-3.0 or a commercial licence, at your option; see LICENSE and LICENSE-COMMERCIAL. The commercial licence is sold yearly and self-service at https://pdfluent.com/sdk/pricing.