HTML to PDF in Python

Updated July 2026 · HTTP client: requests (pip install requests)

The shortest reliable path from HTML to PDF in Python is not a rendering library — it’s an HTTP call to a rendering API. Your app sends HTML, a managed Chromium engine renders it exactly as Chrome would, and a PDF comes back. No browser binaries in your deploy, no native dependencies, no page-break surprises.

Complete example (generate_pdf.py)

Set NIPPYPDF_API_KEY in your environment, then this is the whole program:

# generate_pdf.py — run with: python generate_pdf.py
import os

import requests

resp = requests.post(
    "https://api.nippypdf.com/v1/pdf",
    headers={"Authorization": f"Bearer {os.environ['NIPPYPDF_API_KEY']}"},
    json={
        "html": "<h1>Invoice #1042</h1><p>Total: $2,508.00</p>",
        "format": "A4",
        "margin": "20mm",
        "print_background": True,
    },
    timeout=30,
)
resp.raise_for_status()

with open("invoice.pdf", "wb") as f:
    f.write(resp.content)
print("Wrote invoice.pdf")

All request options

POST https://api.nippypdf.com/v1/pdf accepts JSON; the response body is the binary PDF. Options mirror Chromium’s print pipeline:

{
  "html": "<h1>…</h1>",              // or "url": "https://example.com/invoice/1042"
  "format": "A4",                     // "A4" | "Letter" | "Legal" | { "width": "210mm", "height": "297mm" }
  "margin": "20mm",                   // uniform, or { "top": "…", "right": "…", "bottom": "…", "left": "…" }
  "landscape": false,
  "scale": 1.0,                       // 0.1 – 2.0
  "print_background": true,
  "wait_until": "networkidle",        // "load" | "domcontentloaded" | "networkidle"
  "header_html": "<span></span>",     // running header (use sparingly)
  "footer_html": "<span class=\"pageNumber\"></span> / <span class=\"totalPages\"></span>"
}

For pagination control — keeping blocks together, repeating table headers, paper sizes — see the page-break CSS guide.

The self-hosted alternatives, honestly

• WeasyPrint: a genuinely good pure-Python renderer, but its own engine — output differs from Chrome, and system dependencies (Pango/Cairo) complicate deploys.
• Playwright / pyppeteer: Chromium output, browser-fleet problems — memory, processes, cold starts. Read more →

If those trade-offs are fine for your workload, they are honest choices. The API exists for teams who want Chrome’s output without operating Chrome.

FAQ

Try your own markup in the live in-browser demo, or jump to another language: Node.js, PHP, Ruby, C#, Go.