I would like to install a command line tool within a Docker image in order to quickly convert *html files into *pdf files.

I am surprised there is not a Unix tool to do something like this.

  • 2
    @muru It's arguable a duplicate, though (A) I'm looking for a command line tool to put in a Docker image and (B) the answers below are quite useful and more helpful that the posting above from 2015. I've edited the question to clarify this somewhat, and I'm happy to edit again.
    – EB2127
    Aug 5, 2019 at 4:26
  • 2
    Yes, this question is focused on command line tools while the other isn't and also, the other requires a more complex solution since it's about converting multiple, linked html documents. I don't think it's a dupe.
    – terdon
    Aug 5, 2019 at 16:21
  • 1
    – Barmar
    Aug 5, 2019 at 21:06
  • 1
    this should probably be on Software Recommendations
    – phuclv
    Aug 6, 2019 at 16:47
  • @phuclv This is a good point. I didn't know this existed.
    – EB2127
    Aug 6, 2019 at 22:20

6 Answers 6


pandoc is a great command-line tool for file format conversion.

The disadvantage is for PDF output, you’ll need LaTeX. The usage is

pandoc test.html -t latex -o test.pdf

If you don't have LaTeX installed, then I recommend htmldoc.

Cited from Creating a PDF

By default, pandoc will use LaTeX to create the PDF, which requires that a LaTeX engine be installed.

Alternatively, pandoc can use ConTeXt, pdfroff, or any of the following HTML/CSS-to-PDF-engines, to create a PDF: wkhtmltopdf, weasyprint or prince. To do this, specify an output file with a .pdf extension, as before, but add the --pdf-engine option or -t context, -t html, or -t ms to the command line (-t html defaults to --pdf-engine=wkhtmltopdf).

  • 11
    +1. pandoc can also use wkhtmltopdf to directly convert from html to pdf, without needing latex. see man pandoc and search for wkhtmltopdf or --pdf-engine
    – cas
    Aug 5, 2019 at 4:15
  • 1
    @cas This is really useful. Could you answer the question with that command? I would like to keep this answer
    – EB2127
    Aug 5, 2019 at 4:28
  • @EB2127 Stack Exchange answers can easily contain more than one solution to a problem; collaborative editing can/should make any answer better.
    – Jeff Schaller
    Aug 6, 2019 at 11:05
  • @cas Unfortunately wkhtmltopdf complains about QXcbConnection: Could not connect to display localhost:12.0 and dumps core. I suspect if I figure out the display issue, then it will work but not sure why it cares about the display.
    – steveb
    Mar 12, 2020 at 23:19
  • What advantage is there to using pandoc with the WeasyPrint engine vs just using WeasyPrint without the dependency on pandoc? Jun 4, 2020 at 23:12

You can also try wkhtmltopdf, usage and installation is pretty straightforward.


weasyprint is an option. A possible drawback is that you'll need python on your machine.


pip install weasyprint


weasyprint in.html out.pdf
  • 2
    All distributions are shipped with Python .
    – Paradox
    Aug 6, 2019 at 3:57
  • 4
    Sure, but there are custom linux systems, on embedded devices for example, that might not have python.
    – shiftas
    Aug 7, 2019 at 6:42
  • Tried it, but it ignores # in url. e.g. "status.aws.amazon.com/#AP_Block" converts the wrong tab to pdf
    – Pieter
    Aug 22, 2021 at 10:15

I've been successfully using the 1.8 branch of HTMLDOC for years. I put it in a commercial system that has since generated hundreds of thousands of reports since 2003.

It's not super-versatile, but it is very efficient and reliable. It's limited to a basic set of postscript fonts.

It does not support CSS, but instead uses a special HTML comment directive set to control PDF specific aspects.

The source code is not too difficult to read and edit if you need to add custom facilities, if you're comfortable with C. It is compiled with GCC or Visual Studio, depending on your target platform.

Note that the HTML does not need to be in a file. You can generate it dynamically from a URL, php or aspx etc. You can also hook it up in your web server for generate a PDF file dynamically.

In my use case it generates a PDF file from an asp page which then gets attached to an email, instead of sending the HTML to the printer and the letter stuffing machine; it's a kind of print spooler.

  • 1
    Indeed, a small and usefull tool, with lots of features. Thank you for sharing!
    – Andrei B
    Apr 29, 2020 at 11:54
  • Absolutely useful tool HTMLDOC, we have done some custom changes and have been using it for almost 2 decades. The main issue with HTMLDOC is not supporting CSS and SVG. I tried many open-source options but had no success, at one point in time WKHTMLTOPDF was promising but now the project is archived in 2020, and there are more than 1.3K issues open. I think good dev support will be required to revive this project. I still believe that WK can be a very good substitute for HTMLDOC. Thanks!
    – ppant
    Apr 5, 2023 at 11:38

There is also an html2ps program, and you could then easily convert the PostScript file to pdf. I used this several years ago, and IIRC it did a pretty good job on a large manual.


PhantomJS can do the job for you. It has command line functionality and works out of the box. You'll be required to write a simple Javascript function to tell it what to do. The site has a quick start guide and there are plenty of articles online to assist you. Usage is generally as follows:

phantomjs configFile.js htmlFile.html output.pdf

Here is a sample script to generate an A4 portrait PDF taken from here, save as your configFile.js

var page = require('webpage').create(),
    system = require('system'),
    fs = require('fs');

page.paperSize = {
    format: 'A4',
    orientation: 'portrait',
    margin: {
        top: "1.5cm",
        bottom: "1cm"
    footer: {
        height: "1cm",
        contents: phantom.callback(function (pageNum, numPages) {
            return '' +
                '<div style="margin: 0 1cm 0 1cm; font-size: 0.65em">' +
                '   <div style="color: #888; padding:20px 20px 0 10px; border-top: 1px solid #ccc;">' +
                '       <span>REPORT FOOTER</span> ' +
                '       <span style="float:right">' + pageNum + ' / ' + numPages + '</span>' +
                '   </div>' +

page.settings.dpi = "96";

page.content = fs.read(system.args[1]);

var output = system.args[2];

window.setTimeout(function () {
    page.render(output, {format: 'pdf'});
}, 2000);
  • No longer in development.
    – liftarn
    Nov 21, 2023 at 14:36

Not the answer you're looking for? Browse other questions tagged .