wkhtmltopdf Alternative

wkhtmltopdf has been deprecated since January 2023 with unpatched CVEs and no active maintainers. PDFend is a modern, API-first replacement — same input (HTML + CSS), better output, ten-minute migration.

If you landed here, you probably already know the story: the wkhtmltopdf project was archived on GitHub in January 2023 because its QtWebKit dependency is unmaintainable. QtWebKit itself was dropped by the Qt team in 2016. The result is a binary that embeds an abandoned browser engine with multiple open CVEs, no Flexbox, no Grid, and memory leaks that get worse over time. Teams keeping wkhtmltopdf in production today are doing so with crossed fingers.

PDFend was built to be the obvious migration target. Same mental model — you send HTML, you get a PDF — but the rendering happens on current Chromium, running in our managed infrastructure, reachable with a single HTTP POST.

The short version: wkhtmltopdf was best-in-class in 2015 and has not meaningfully changed since. PDFend is what you would build if you started today.

Most migrations are a one-function change: replace the shell-out to wkhtmltopdf with an HTTP POST, keep your existing HTML templates, verify the output. We'll show the common stacks side-by-side.

1. Get a free API key

Sign up at pdfend.com/sign-up, create an API key in the dashboard. No credit card required.

2. Replace your wkhtmltopdf call

Node.js

// Before — spawning wkhtmltopdf as a subprocess
spawnCliTool("wkhtmltopdf", ["-", "output.pdf"], html);
const pdf = await fs.readFile("output.pdf");

// After — PDFend
import { PDFend } from "@pdfend/sdk";
const pdfend = new PDFend(process.env.PDFEND_API_KEY!);
const { url } = await pdfend.generate({ html });

Python (Django)

# Before — pdfkit (wkhtmltopdf wrapper)
import pdfkit
pdf = pdfkit.from_string(html, False)

# After — PDFend
from pdfend import PDFend
client = PDFend(os.environ["PDFEND_API_KEY"])
result = client.generate(html=html)
pdf_url = result["url"]

Ruby (Rails)

# Before — wicked_pdf
pdf = WickedPdf.new.pdf_from_string(html)

# After — PDFend via net/http
uri = URI("https://api.pdfend.com/v1/generate")
req = Net::HTTP::Post.new(uri,
  "Authorization" => "Bearer #{ENV['PDFEND_API_KEY']}",
  "Content-Type" => "application/json")
req.body = { html: html }.to_json
res = Net::HTTP.start(uri.hostname, uri.port, use_ssl: true) { |h| h.request(req) }
pdf_url = JSON.parse(res.body)["url"]

PHP (Laravel)

// Before — snappy/wkhtmltopdf
$pdf = PDF::loadHTML($html)->output();

// After — PDFend via Guzzle
$res = Http::withToken(env('PDFEND_API_KEY'))
    ->post('https://api.pdfend.com/v1/generate', ['html' => $html]);
$pdfUrl = $res->json('url');

3. Download or redirect

PDFend returns a signed URL. You can download the PDF server-side and attach it to an email, redirect the user, or store the URL in your database.

A handful of wkhtmltopdf behaviors differ from browsers enough that pixel-exact migrations need attention. None are blockers.

• Inline-block margin rounding. wkhtmltopdf rounded sub-pixel margins differently from Chromium. A layout that looked perfect in wkhtmltopdf may shift by 1-2px in the migrated version.
• --disable-smart-shrinking.wkhtmltopdf's “smart” content-aware scaling is off by default on PDFend. Use options.scale to adjust.
• Header/footer HTML passed via files. wkhtmltopdf expected separate HTML files; PDFend takes them inline as strings. Inline your templates into the request body.
• JavaScript timing. PDFend waits for network idle before printing, so async charts render correctly — but if your page has an infinite polling loop, it will hit the 30-second sync timeout. Use DOMContentLoaded or skip polling for print mode.