Getting Started with PDF Oxide CLI

pdf-oxide is a command-line tool for fast, local PDF processing. It ships the same Rust core that powers the library — zero cloud, zero dependencies.

Quick Install (Recommended)

One command — auto-detects your OS and architecture:

Linux / macOS

curl -fsSL oxide.fyi/install.sh | sh

Windows (PowerShell)

irm oxide.fyi/install.ps1 | iex

Alternative Installation

Homebrew (macOS / Linux)

brew install yfedoseev/tap/pdf-oxide

Cargo

cargo install pdf_oxide_cli

cargo-binstall (pre-built binary)

cargo binstall pdf_oxide_cli

Pre-built binaries

Download from GitHub Releases and add to your PATH.

Quick Start

# Extract text
pdf-oxide text document.pdf

# Convert to Markdown
pdf-oxide markdown document.pdf -o output.md

# Convert to HTML
pdf-oxide html document.pdf -o output.html

# Show PDF info (page count, metadata, encryption)
pdf-oxide info document.pdf

# Search for text (regex supported)
pdf-oxide search document.pdf "pattern"

Commands

Command	Description	Key flags
text	Extract plain text	--format plain|words|lines, --area x,y,w,h
paths	Extract vector paths	--format json|rects|lines, --area x,y,w,h
markdown	Convert to Markdown	—
html	Convert to HTML	—
info	Show metadata and page count	—
search	Search for text (regex)	-i (ignore case), --pages
images	Extract images	--area x,y,w,h, --pages
render	Render pages to PNG/JPEG	--dpi N, --format png|jpeg, --quality N
merge	Merge multiple PDFs	(2+ files required)
split	Split into individual pages	—
create	Create PDF from Markdown / HTML / text	--from markdown|html|text
compress	Compress and optimize	—
encrypt	Encrypt with password (planned v0.4.0 — placeholder today)	—
decrypt	Decrypt a protected PDF	--password PW
rotate	Rotate pages	--degrees 90|180|270|-90
delete	Remove pages	--pages
reorder	Reorder pages	--order "3,1,2,5,4"
metadata	Read, edit, or strip metadata	--title, --author, --subject, --keywords, --strip
watermark	Add text watermark	--opacity 0.3, --rotation 45, --font-size 48, --color R,G,B
bookmarks	List document outline	—
flatten	Flatten annotations or forms	--forms, --annotations
crop	Crop page margins	--margins L,R,T,B
forms	List, fill, or export form fields	--fill "k=v,…", --export fdf|xfdf, --area x,y,w,h

23 commands total. Run pdf-oxide <command> --help for full flag documentation on any command.

Global Options

-o, --output <PATH>        Output file (defaults to stdout for text outputs)
-p, --pages <RANGE>        Page range: "1-5", "1,3,7", "1-3,7,10-12"
    --password <PW>        Password for encrypted PDFs
-v, --verbose              Show timing information
-q, --quiet                Suppress non-essential output
-j, --json                 Output as JSON
    --no-banner            Skip the REPL banner

Examples

Extract pages 1–5 as Markdown:

pdf-oxide markdown report.pdf --pages 1-5 -o report.md

Merge three PDFs:

pdf-oxide merge cover.pdf body.pdf appendix.pdf -o combined.pdf

Add a watermark:

pdf-oxide watermark contract.pdf "DRAFT" --opacity 0.2 --font-size 72

Fill form fields:

pdf-oxide forms application.pdf --fill "name=Jane Doe,date=2026-02-28" -o filled.pdf

Search with case-insensitive regex:

pdf-oxide search paper.pdf "neural.?network" -i

Render page 1 as a 300-DPI PNG:

pdf-oxide render report.pdf --dpi 300 --format png --pages 1 -o page1.png

Extract the vector path overlay from a schematic as JSON:

pdf-oxide paths schematic.pdf --format json -o paths.json

Interactive REPL

Run pdf-oxide with no arguments to enter interactive mode:

$ pdf-oxide
pdf-oxide> text /path/to/doc.pdf
(extracted text appears)
pdf-oxide> info /path/to/doc.pdf
(metadata appears)