Gen-PDF Review (2026): Pricing, Features & Honest Verdict

TLDR

Gen-PDF is an MCP server and API that converts markdown into clean, formatted PDFs using the Typst engine. It does one job well: markdown in, publication-ready PDF out, straight from an AI agent or script. Best for: developers and AI agents automating document generation. Price: Free 20 pages/mo; Pro $9/mo. Rating: 6.8/10.

What is Gen-PDF?

Gen-PDF is a tool for turning markdown into professionally formatted PDF documents. It ships as a Model Context Protocol (MCP) server, which means an AI assistant like Claude or an agent running in Cursor can call it directly to produce a PDF. There is also a hosted API at mcp.gen-pdf.com for developers who want to wire it into their own code. Under the hood it uses Typst, a modern Rust-based typesetting engine, instead of the older LaTeX or headless-browser approaches most PDF tools rely on.

We dug into it because PDF generation is one of those boring problems that AI workflows keep bumping into. An agent writes a report, drafts an invoice, or compiles technical documentation, and then someone has to turn that text into a real document. Historically that meant either a clunky headless-browser pipeline, a LaTeX setup that takes a day to configure, or a copy-paste into Word that breaks every code block. Gen-PDF aims at exactly that gap. It is a young, focused product distributed through Smithery and GitHub, and it is built for the agent era rather than as a general consumer app. That focus is both its strength and the reason it feels more like a utility than a platform.

The reason it matters now is the rise of agentic workflows. As more teams hand off real tasks to AI agents, the agents need to produce real artifacts, and a PDF is one of the most common. A tool that an agent can call mid-task to emit a finished document removes a manual step that used to need a human. That is the bet Gen-PDF is making, and it is a sensible one.

What Are Gen-PDF’s Key Features?

Markdown to PDF via Typst

The core feature is a single generate_pdf tool that takes markdown and returns a styled PDF. Because it runs on Typst rather than spinning up a browser, rendering is fast, often in milliseconds for short documents. You get cover pages, automatic table of contents, and complex tables including cell merging, which is more than most lightweight converters handle.

Rich content support

Gen-PDF handles the things technical writers actually need: LaTeX mathematical expressions, Mermaid diagrams and flowcharts, syntax highlighting across more than 100 programming languages, and embedded images and media. If you are generating engineering docs or data analysis write-ups, these are the details that usually break cheaper tools.

MCP and API integration

Setup is the selling point. You can install it for Claude Desktop with a single Smithery command, or connect to the hosted MCP endpoint using an API key from the Gen-PDF dashboard. That means an agent can generate a finished PDF mid-conversation with no manual export step. For developers, the same endpoint works as a normal API call inside any pipeline.

Styling controls

You get customizable styling including dark mode, typography controls, and document formatting options. It is not a drag-and-drop template builder, but for a tool driven by code or an agent, having styling knobs you can pass as parameters is more useful than a visual editor you cannot automate. You can tune fonts, spacing, and layout through the generation call, which keeps everything reproducible. If you generate the same invoice template a thousand times, you want it defined in code, not clicked together by hand each run.

Speed from the Typst engine

The choice of Typst over LaTeX or a headless browser is the technical detail that pays off. Typst is written in Rust and compiles fast, so short documents render in milliseconds rather than the seconds a Chromium-based renderer typically needs to spin up. For an agent generating documents in a loop, or an API serving many requests, that latency difference compounds. It also means fewer moving parts than a browser-based stack, which tends to reduce the weird font and rendering bugs those pipelines are known for.

How Much Does Gen-PDF Cost?

Gen-PDF starts free. The free tier gives you 20 PDF pages per month with every feature unlocked and no credit card required, which is genuinely enough to evaluate it or power a small side project. The Pro plan is $9 per month and includes 1,000 pages, then bills $0.001 per page after that, with no watermark and priority support. There is a 30-day money-back guarantee.

Against commercial alternatives, that is cheap. DocRaptor, a mature HTML-to-PDF API, starts at $15/mo for 125 documents and climbs to $75/mo for 1,000. PDFShift starts around $9/mo but meters by output file size in 5MB credits. Api2Pdf is the closest on price, charging roughly $0.001 per PDF pay-as-you-go. Gen-PDF lands in that pay-as-you-go bracket but bundles a free monthly allowance most rivals do not.

One caution: the billing unit is pages, not documents. A long, image-heavy report can eat your allowance faster than a per-document plan would. For typical short reports and invoices the math works out well, but if you generate dense documents at scale, model your real page counts before committing.

Who is Gen-PDF Best For?

Use Gen-PDF if you are a developer or building AI agents and you need formatted PDFs to come out of an automated workflow. It fits solo makers, small teams, and agencies generating reports, invoices, contracts, or technical documentation where someone, or something, is producing markdown already. The MCP integration is the killer detail: if your stack runs through Claude or Cursor, this slots in with almost no friction.

Skip Gen-PDF if you want an interactive document editor, a library of visual templates, or document storage and collaboration. It has none of that and is not trying to. It is also not the pick if you only generate a handful of PDFs a month by hand, since a free browser print-to-PDF does that job. This is a tool for automation, not for occasional manual exports.

It is also worth thinking about whether you want a hosted dependency at all. If your documents contain sensitive data and your security posture rules out sending content to a third-party endpoint, the open-source self-hosted converters are a better fit even though they cost you setup time. Gen-PDF is the convenient choice when speed of integration outweighs the desire to keep everything in-house, which describes most small teams and indie builders but not every enterprise.

Best Gen-PDF Alternatives

markdown2pdf MCP (open source)

If you are comfortable self-hosting, the open-source markdown2pdf and Pandoc MCP servers do the same basic job for free. You install them locally, point them at pandoc or weasyprint, and your only cost is the machine running them. The tradeoff is setup and maintenance: no hosted endpoint, no support, and you handle scaling yourself. Worth it for tinkerers and teams that want full control.

DocRaptor

DocRaptor is the established commercial option, an HTML-to-PDF API that has been around for years. It starts at $15/mo for 125 documents and scales to $299/mo for 5,000. It is more proven on reliability and support than Gen-PDF, and it works from HTML rather than markdown, which suits teams already templating in HTML. You pay more for that maturity.

Api2Pdf

Api2Pdf is the bare-metal price play: pure pay-as-you-go at roughly $0.001 per PDF with no monthly minimum to speak of. It is a lower-level API without the markdown and Typst niceties, so you bring your own formatting. Pick it if you want the cheapest possible per-document cost and do not need the MCP integration or rich markdown rendering.

Final Verdict: Is Gen-PDF Worth It?

Gen-PDF is a sharp, well-scoped utility. It turns markdown into clean PDFs, runs fast thanks to Typst, supports the math, diagrams, and code highlighting that technical documents actually use, and drops into MCP clients with one command. At free for 20 pages and $9/mo for 1,000, the price is fair and undercuts most commercial PDF APIs. If your workflow already involves an agent or code emitting markdown, this is an easy add.

The reasons to hesitate are its youth and its narrow scope. It is new and small, so there is little public track record on uptime or support, and it does exactly one thing with no editor, templates UI, or storage. The page-based metering also needs watching on long documents. We landed on 6.8 out of 10: a genuinely useful tool that does its single job well, held back from a higher score only by how unproven and limited in scope it still is. Recommended for the right automated use case, not as a general document tool.