html2pdf-slides

Convert HTML slide presentations to high-fidelity PDF files. The fastest way to export reveal.js, Slidev, Marp, impress.js, and Quarto presentations as PDF, with animations, fragments, dark themes, and click-reveal content fully preserved.

TL;DR

html2pdf-slides is an open source command-line tool and Node.js library that converts HTML slide presentations to PDF. It auto-detects reveal.js, Slidev, Marp, impress.js, Quarto, and Shower decks, walks through every slide and animation state via framework-native APIs, and assembles the screenshots into a pixel-perfect PDF with correct dimensions and per-slide background colors. Built on Puppeteer, zero configuration, MIT licensed.

A modern, framework-aware HTML to PDF converter for slide decks. Works with reveal.js, Slidev, Marp, impress.js, Quarto, Shower, custom HTML slides, and any HTML presentation that uses arrow-key navigation.

Quick Start

# Install globally
npm install -g html2pdf-slides

# Or use without installing
npx html2pdf-slides https://revealjs.com/demo/ -o slides.pdf

# Local file
html2pdf-slides presentation.html -o output.pdf

# Online reveal.js presentation
html2pdf-slides https://revealjs.com/demo/ -o revealjs-demo.pdf

# Slidev deck (from deployed URL, no source code needed)
html2pdf-slides https://your-slidev-deck.netlify.app/ -o deck.pdf

# Marp presentation
html2pdf-slides https://yhatt.github.io/marp-cli-example/ -o marp.pdf

Why html2pdf-slides?

Most HTML to PDF tools treat your presentation like a static webpage. They miss animations, break dark themes, and collapse slide layouts. Generic HTML to PDF converters such as wkhtmltopdf, Puppeteer's page.pdf(), or browser print dialogs were never designed for slide decks.

html2pdf-slides uses framework-native APIs to navigate through every slide and fragment state, captures pixel-perfect screenshots, and assembles them into a clean PDF with correct dimensions and background colors. It is the easiest way to export presentation slides as PDF without losing visual fidelity.

Common use cases

• Export reveal.js presentation to PDF for distribution or printing
• Convert Slidev deck to PDF without rebuilding from source
• Save Marp slides as PDF preserving themes and code highlighting
• Archive impress.js presentations as a static PDF document
• Generate PDF handouts from Quarto reveal.js slides
• Batch convert HTML slide decks to PDF in CI/CD pipelines
• Capture deployed presentations from any URL, no local source needed

Supported Frameworks

| Framework | How it captures | |-----------|----------------| | reveal.js | Uses Reveal.next() to walk all slides, vertical slides, and fragment steps | | Slidev | URL-based navigation with full v-click content reveal (all 8 hidden states) | | Marp | Hash navigation via bespoke.js | | impress.js | impress().goto() API for 3D-transformed steps | | Quarto reveal.js | Waits for async Reveal.js initialization, then uses Reveal API | | Custom HTML (.slide class) | html2canvas capture, auto-discovers all .slide elements | | Shower | Same as above, detected via .slide class | | Any other HTML slides | Keyboard fallback (ArrowRight, Space, PageDown) with screenshot deduplication |

Comparison

| | html2pdf-slides | DeckTape | Browser Print | Native export | |---|---|---|---|---| | Setup | Auto-detect, zero config | Needs --plugins flag | Manual | Needs source code | | Dark themes | Preserved | Preserved | Stripped to white | Varies | | Fragments | One page per state | Configurable (separate or overlay) | Final state only | Varies | | Slidev v-click | All 8 states | Partial | No | Yes | | Slide backgrounds | Per-slide sampling | Single color | White | Varies | | Quarto support | Yes (async wait) | May fail | No | N/A | | Fallback for unknowns | Multi-key + dedup | Single method | N/A | N/A |

Native export refers to built-in tools like reveal.js ?print-pdf, slidev export, or marp-cli. These require project source code and local dev setup. html2pdf-slides works from any deployed URL.

How It Works

1. Detect the slide framework from DOM markers
2. Navigate using framework-native APIs (not generic scrolling)
3. Capture pixel-perfect screenshots at original viewport size
4. Assemble into PDF with per-slide background colors and 16:9 minimum page dimensions

Vector mode (lossless, selectable text)

By default html2pdf-slides captures slides as high-resolution screenshots (raster mode), which preserves every pixel including custom fonts, gradients, and canvas-rendered content. For decks that are mostly text and SVG, --mode vector produces a smaller PDF where text stays selectable and searchable and graphics stay crisp at any zoom level.

# Raster mode (default): pixel-perfect, larger file, text is not selectable
html2pdf-slides slides.html -o slides.pdf

# Vector mode: smaller file, selectable text, scalable SVG
html2pdf-slides slides.html --mode vector -o slides.pdf

When to pick which:

| Use case | Mode | |---|---| | Code-heavy decks, want copy-pasteable code blocks | vector | | Text-heavy presentations, want full-text search in the PDF | vector | | Canvas / WebGL / complex CSS animations | raster (default) | | Custom fonts that may not embed cleanly | raster (default) | | Smallest possible file size | vector |

Vector mode uses Chromium's native page.pdf() per slide and merges with pdf-lib, so text-as-text and SVG-as-vector are preserved.

Framework support in vector mode (as of 1.1.0): generic .slide-class decks (with or without an active class). Framework-specific decks (reveal.js, Slidev, Marp, impress.js) are auto-detected and rejected with a clear error, because they need framework-aware navigation, which lands in 1.2.0. For those, use raster mode (the default).

Note on --page-width in vector mode: --page-width controls PDF page width (in points) in raster mode. Vector mode ignores it and emits pages at the source viewport dimensions to preserve the deck's natural aspect ratio without scaling. If you need a specific PDF page width, use raster mode.

Options

| Flag | Description | Default | |------|-------------|---------| | -o, --output <path> | Output PDF path | same as input | | -s, --selector <css> | CSS selector for slides | .slide | | -b, --bg-color <hex> | Background color | #0a0a0a | | --mode <raster\|vector> | Capture mode (see Vector mode section above) | raster | | --scale <n> | Resolution multiplier (raster mode) | 1.5 | | --quality <n> | JPEG quality 1-100 (raster mode) | 88 | | --page-width <pts> | PDF page width | 842 | | --parallel <n> | Parallel browser tabs | 2 | | --wait <ms> | Wait before capture | 300 | | --retry <n> | Retry for blank slides | 2 | | --no-headless | Show browser window | |

Programmatic API

import { convertToPDF } from 'html2pdf-slides';

const result = await convertToPDF({
  input: 'https://revealjs.com/demo/',
  output: 'slides.pdf',
  quality: 90,
  mode: 'raster', // or 'vector' for selectable-text output
  onProgress: (current, total) => {
    console.log(`Capturing slide ${current}/${total}`);
  },
});

// result: { slideCount, blankCount, pdfSize, outputPath, framework }

FAQ

How do I convert reveal.js slides to PDF?

html2pdf-slides https://your-reveal-presentation.com -o output.pdf

Automatically detects reveal.js and uses the Reveal API to navigate through all slides and fragments. Preserves your theme, animation final states, and dark backgrounds.

How do I convert Slidev presentations to PDF?

html2pdf-slides https://your-slidev-deck.netlify.app -o output.pdf

Converts any deployed Slidev presentation directly from its URL. No source code needed. Fully handles v-click content visibility.

Can I convert any HTML presentation to PDF?

Yes. For unrecognized frameworks, it falls back to keyboard navigation with screenshot deduplication to detect when all slides have been captured.

Why are some slides blank in the output?

Try increasing the wait time for slides that rely on lazy loading:

html2pdf-slides presentation.html --wait 1000 -o output.pdf

How is html2pdf-slides different from DeckTape?

DeckTape is the long-standing tool for this task but requires manual --plugins flags per framework, lacks Quarto async support, and only captures partial Slidev v-click states. html2pdf-slides auto-detects the framework, handles all 8 Slidev v-click hidden states, supports Quarto's async Reveal initialization, and ships a multi-key keyboard fallback for unknown frameworks.

Can I use html2pdf-slides as a library in my Node.js project?

Yes. Install as a dependency and import convertToPDF from the package. See the Programmatic API section above for an example. Useful for generating PDF handouts inside CI/CD pipelines or build scripts.

Does html2pdf-slides work with Marp slides?

Yes. It detects Marp's bespoke.js runtime and navigates via hash-based URL changes. Both marp-cli HTML output and Marp for VS Code exports are supported.

How do I export a Quarto reveal.js presentation to PDF?

Pass the deployed Quarto site URL directly. html2pdf-slides waits for Quarto's async Reveal.js initialization (which DeckTape often misses), then captures every slide.

html2pdf-slides https://your-quarto-site.com/slides.html -o quarto.pdf

Requirements

• Node.js 18 or later
• Chromium is downloaded automatically by Puppeteer on first run

License

MIT