Gotenberg Alternative: Managed HTML to PDF With No Container to Run

If you can run a container and you want to, Gotenberg is hard to beat. It is free, open source and actively maintained, it renders HTML and URLs with real Chromium so the output matches a browser, and it adds something a plain renderer does not: LibreOffice backed conversion of Word, Excel and PowerPoint files to PDF, plus merge, split, watermark and Factur-X e-invoicing. Because it runs on your own infrastructure, your documents never leave your network, which matters when data residency or a compliance boundary rules out sending pages to a third party. For those requirements, self hosting Gotenberg is the correct decision and you should not pay anyone to replace it.

The question this page answers is narrower: what do you do when you want that Chromium quality output but you do not want to be the team that runs and scales the browser? That is where a hosted API earns its place.

Gotenberg hides the browser behind a tidy API, but it does not remove it. In production you are operating headless Chromium, and that carries the same bill as any browser service. A single Chromium instance handles only about six parallel operations, and a single LibreOffice instance cannot run conversions in parallel at all because of its lock, so throughput is capped per container. To go faster you run several containers behind a load balancer, usually orchestrated with Kubernetes or ECS, and you allocate 512MB to 1GB of RAM for each Chromium instance. Then you add the parts nobody enjoys: monitoring, alerting, Docker image updates, and handling the moment Chromium runs out of memory and crashes under a burst of traffic.

None of this is a knock on Gotenberg; it is the nature of rendering in a browser. But it is real engineering time and real hosting spend, and for many teams the PDF feature is not worth standing up and babysitting a browser fleet. The library versus API comparison puts numbers on that tradeoff, and the Puppeteer alternative page covers the same self host versus hosted decision from the raw browser angle.

Sitepdf runs the same managed Chromium that Gotenberg would run for you, but on our infrastructure. You send a URL or raw HTML in one HTTP request and get back a pixel faithful PDF: your CSS, web fonts and JavaScript all honored, because a real browser does the rendering. There is no container to deploy, no concurrency cap you have to plan around, no RAM to provision, and nothing to patch. When a Monday morning batch spikes your volume, that is our scaling problem, not a pod count you tuned last quarter.

The second difference is the archive. Add archive=true to the render call and Sitepdf stores a timestamped snapshot of exactly what was rendered, retrievable later by id. Gotenberg generates a file and hands it back; there is no built in dated record of the source. If you ever need to prove what a document looked like when you produced it, for a dispute, an audit, or a compliance requirement, that website archiving capability is the part you cannot easily bolt onto a self hosted renderer.

Honesty first: if you need to convert Word, Excel or PowerPoint files to PDF, Gotenberg does that through LibreOffice and Sitepdf does not. We render HTML and URLs to PDF, and we archive them; we are a document grade web renderer, not an Office conversion suite. If your workload is mostly turning uploaded .docx and .xlsx files into PDFs, keep Gotenberg or use a LibreOffice based service. If your workload is turning your own HTML templates and live web pages into faithful, dated PDFs at scale without operating a browser, that is exactly what Sitepdf is for. The two even coexist well: some teams keep Gotenberg for Office files and call a PDF generator API for their high volume HTML rendering.