100% Pure-Rust Engine • Zero C++ DLLs • Offline AI OCR

The Lightning-Fast PDF Reader
Built in Pure Rust

PDFbull pairs a high-performance document rendering pipeline (zpdf) with native neural OCR (rten), PKCS#12 digital signatures, geospatial GIS analysis, and 1-click Markdown / HTML5 export.

Install via Cargo / Windows PowerShell
$ cargo install pdfbull
sample_document.pdf — PDFbull
PDFbull UI Showcase

Engineered for Extreme Speed & Privacy

Unlike bloated PDF readers tied to legacy C++ binaries (PDFium/MuPDF), PDFbull is engineered natively in Rust for instant startup, low memory footprint, and 100% offline security.

🧠

Offline AI OCR Suite

Built-in ocrs + rten neural engine. Detects and extracts text from scanned pages offline out-of-the-box with English/Latin and Devanagari script models.

Devanagari OCR Latin / English 0 Network Calls
🔒

PKCS#12 Digital Signatures

Cryptographically sign PDF documents with standard .p12 / .pfx certificates, verify signature trust chains, and overlay rubber stamps (APPROVED, CONFIDENTIAL, DRAFT).

PKCS#12 / PKCS#8 Visual Stamps Trust Chain
📄

Blank PDF Creator & Page Grid

Create new blank PDFs from scratch with DocumentBuilder, rotate pages, reorder visually in a grid, merge multiple PDFs, or split documents seamlessly.

DocumentBuilder Merge & Split Page Organizer
🔄

1-Click Markdown & HTML5 Export

Transform raw PDF content into structured Markdown (.md), semantic HTML5 (.html), or clean text (.txt) with full table and line hierarchy preservation.

Markdown .md HTML5 .html Plain Text .txt
🗺️

Geospatial GIS & CMYK Inspector

Inspect GeoPDF /Measure dictionaries for GIS coordinate systems and inspect embedded CMYK ↔ RGB subtractive color space metrics live.

GeoPDF /Measure CMYK Color Inspector PDF/A Validator

Pure-Rust Multi-Script OCR Engine

PDFbull bundles rten neural models directly to recognize scanned document pages without sending a single byte to cloud servers.

नमस्ते संसार! PDFbull ऑप्टिकल कैरेक्टर रिकग्निशन इंजन।
Detection Model text-detection.rten (DBNet PP-OCRv4)
Recognition Model devanagari_PP-OCRv4_rec.rten
Runtime Latency < 45ms per page (Pure CPU SIMD)
Network Calls 0 (100% Offline Standalone)

Benchmark Comparison

Compare PDFbull's resource utilization against mainstream legacy PDF software.

Feature / Metric PDFbull (v0.19.3) SumatraPDF Chrome (PDFium) Adobe Acrobat Evince
Core Engine Language Pure Rust (zpdf) C / C++ (MuPDF) C++ (PDFium) C / C++ C (Poppler)
120 MB PDF Open Time ~79 ms ~385 ms ~692 ms ~681 ms ~850 ms
Page Render Throughput 1,260 FPS ~30 FPS ~22 FPS ~15 FPS ~20 FPS
Built-in Offline OCR Yes (rten + ocrs) No No Paid Cloud / DLL No
Devanagari Script OCR Yes (Out-of-the-box) No No Complex Setup No
1-Click Markdown / HTML Export Yes (Native GUI) No No Paid Subscription No
C++ DLL Dependencies 0 (Pure Native) Requires DLL Multi-Process 50+ DLLs 15+ Shared Libs
PDFbull Scorecard & Metrics

Download PDFbull for Your OS

Free, open-source under MIT & Apache-2.0 licenses. A Windows installer is built for every release; the macOS and Linux builds were last published with v0.15.1.

🪟

Windows 10 / 11

64-bit Native Windows Application with NSIS Installer

Version 0.19.3 • x86_64-pc-windows-msvc
🍎

macOS

Separate Intel (x86_64) and Apple Silicon (aarch64) builds

Last published in v0.15.1 • x86_64 / aarch64
🐧

Linux

Standalone x86_64 and aarch64 tarballs

Last published in v0.15.1 • x86_64-unknown-linux-gnu