Apryse: The Only SDK Built for the Entire Document Lifecycle
Quick Comparison
Company
600+ global employees with localized support
100+ global employees with limited regional coverage
Active users
Billions, used by 85% of the Fortune 100 and many popular SaaS products
Millions, claims billions due to Google Chrome using PDFium, which is not their product
Front-end
Open source, modular UI that’s easy to extend
Closed source, only customizable via API
Core
Proprietary, in-house Core SDK (single vendor)
PDFium-based fork (open source)
Rendering complex files
Superior performance with optimizations available via linearization (streaming byte ranges) or flattening
Struggles with large, complex files, optimizations available via linearization or flattening
Office Editing
Native in-browser DOCX editing, XLSX editor available
DocJSON intermediary for authoring, recommends not storing as DOCX as their support “remains best effort”
Conversion
Highly accurate, especially when converting to and from structured formats such as office files
Struggles with complex elements such as page breaks, tables, and list content
Third-party dependencies
Built to run without external runtime dependencies
PDFium-based, cloud required for certain features, third party LLM required for certain features
Trusted by Leaders
Join the ranks of over 20,000 innovative start-ups, governments and Fortune 500 companies - including more than 85% of Fortune 100 companies that trust Apryse
Powering the Full AI Document Lifecycle
Apryse is your strategic partner for the document-to-data lifecycle, empowering developers to operationalize AI strategies before and after the model. From preparing and extracting structured JSON via our reliable core engine to providing a flexible, high-fidelity front end for human review, Apryse handles the heavy lifting across your entire pipeline.
Apryse vs “AI-native” Competitors
UI
Open-source UI that supports your custom LLM hooks
Productized viewer hooks that call an LLM from a closed-source UI
Prompt Log
Fully customizable and can connect to any model, MCP, or RAG workflow
Underlying prompt logic is not reviewable or customizable
Deployment Flexibility
Run Apryse fully in your infrastructure without creating additional dependencies
“Works with our server” features that pull your data through a single path
Data Extraction
ML-trained data extraction finely tuned to popular enterprise use cases; LLM optional
Data extraction workflows that feed your data through an LLM
Nutrient is powered by PDFium
PDFium is an open-source rendering engine built by Foxit/Google, launched via Google/Chromium, and adopted by many vendors. Nutrient has 8 verifiable contributions to the 17,686 total code commits (Oct 2025). By contrast, Apryse customers directly access support from the engineers who built the SDK, and is constantly being improved.
Proprietary technology with no dependencies on cloud hosting or external LLMs, Apryse deploys securely within your application.
Penetration Testing
Annual
Annual
Resolved CVEs
13/13 (November 2025)
?/113 (November 2025)
SOC 2
SOC 2 Type II
SOC 2 Type II
“The performance of the Apryse SDK was excellent in terms of rendering times and editing task completion rates … [and] especially compared to other vendors that Dropbox has worked with in the past, the Apryse engineering support was just very responsive”
Sr Dropbox Product Manager, Amanda Lansman
Feature Chart: See How Apryse is MADE BETTER
Viewing
High performance viewer, side-by-side and multi-tab view modes, excels with complex files
Competent viewer, side-by-side mode, restrictive zoom/memory ceilings, struggles with complex files
Language Localization
30+ localizations, RTL UI and content interactions
29 localizations, does not support RTL content interactions
UI Customization
Open-source modular UI, quick to extend or theme
Closed source, enable/disable elements via API
UI Accessibility
Meets WCAG 2.2 AA guidelines
Meets WCAG 2.2 AA guidelines
Annotations
Extensive OOTB annotations, deep customization, import/export
Basic OOTB annotations, import/export
Search
Pattern search + index panel, SDK performance enhances handling of long docs
Pattern search, no index panel by default
Forms
Create and fill with live editing
Uses placeholder annotations
Signatures
Live editing of created signature fields, works across all platforms
Long-term validation only on mobile / windows
Measurements
6 measurement types, snap-to-point, unit systems
4 measurement types, unit systems
Redaction
Permanent redaction, pattern-based search & redact
Permanent redaction, LLM-assisted search & redact
PDF Content Editing
Edit text & images, deep rich text editing, highly responsive
Edit text, limited rich text editing, can only resize paragraphs horizontally
DOCX Editing
Native DOCX editing (no intermediate format)
DocJSON intermediary, requires DOCX import/export
Spreadsheet Editing
In-browser XLSX editing
N/A
OCR
Modified version of Tesseract or Iris OCR
Proprietary ORPALIS OCR
Text & Image Extraction
Extract plain text & images programmatically
Extract plain text programmatically
Key-Value Pair Extraction
ML-assisted KVP to JSON
ML-assisted KVP to JSON
Tabular Data Extraction
ML-assisted table detection to JSON/XLSX
Basic table extraction to JSON
Document Structure Extraction
Extract headers, paragraphs, lists, tables, etc to JSON
N/A
Document Classification
Page-level document classification with confidence scores
LLM-assisted classification when connected
Barcode Extraction
Barcode decoding with preprocessing for skew/quality
Barcode decoding with preprocessing for skew/quality, only in .NET applications
CAD Title Block Extraction
Extract standard and non-standard title block information from technical drawings to JSON
N/A
Template Generation
Generate via merging JSON data with PDFs, or from Office templates using Fluent
Generate via merging JSON data with PDFs
Image Conversion
~30 image types including advanced (e.g. DICOM)
~11 basic image types
Office Conversion
Industry leading conversion quality to and from Office
Struggles with rich text and structure in Office files, however, can perform entirely client-side
CAD Conversion
6 CAD file types
2 CAD file types
PDF/A Conversion
Create & validate PDF/A-1 to A-4
Create & validate PDF/A-1 to A-4
Auto-Tagging
Adds/fixes tags using Apryse SDK, PDF/UA friendly output via iText
Adds tags, PDF/UA friendly output via costly cloud services
Page Comparison
Side-by-side, overlay, semantic text, and image comparison
Side-by-side, overlay, semantic text, and image comparison
What Our Customers Say
"The speed at which they integrate is superior to their competition and their product roadmap has been good -- they’ve invested in the right things."
Marcus O'Brien
Global Head of Product Management, AutoCAD
"Apryse SDK [PDFTron] had particularly strong annotation and collaboration features, and their cross-platform approach allowed us to accelerate how we delivered products."
Sam Stuart
Senior Product Manager, Bentley Systems
"Apryse [PDFTron] demonstrated superior speed and functionality out of the box. In contrast, the competition took a lot of shortcuts and it was obvious in the user experience."
Alistair Michener
Founder and CEO, Drawboard
Apryse is the top choice for document processing technology.
