No. Texify uses LaTeX-quality document rendering. HTML input may come later, but the current focus is structured data plus LaTeX templates.

Do I need to know LaTeX?

For the beta, yes. Later, templates and examples should reduce the LaTeX knowledge required to get started.

Can I use my own templates?

Yes. The beta supports user-provided LaTeX templates.

Can I send raw LaTeX?

Yes. POST raw LaTeX files as resources to /render and save the application/pdf response body.

Is this safe for private documents?

During beta, avoid sending highly sensitive production documents. Security and isolation are core roadmap items, but we will not claim enterprise security before it exists.

Why not use HTML-to-PDF?

HTML is great for simple documents. Texify targets documents where pagination, tables, math, and layout predictability matter.

PDF template API

Typeset-quality PDF API — LaTeX in, polished PDF out.

Real typesetting engine, deterministic pagination, and professional typography through a REST API, remote MCP server, agent skill, and gallery of LaTeX templates you copy and render. No headless browser.

Try the API

Browse templates →

View sample PDF →

JSON data

{
  "invoiceNumber": "INV-2026-001",
  "customer": {
    "name": "Acme GmbH"
  },
  "items": [
    {
      "description": "Backend development",
      "quantity": 10,
      "unitPrice": 120
    },
    {
      "description": "API integration",
      "quantity": 4,
      "unitPrice": 140
    }
  ],
  "total": 1760
}

LaTeX template

\section*{Invoice [[ .invoiceNumber ]]}

\begin{longtable}{l r r}
  \toprule
  Description & Qty & Unit \\
  \midrule \endhead
  [[ range .items ]]
  [[ .description ]] &
  [[ .quantity ]] &
  [[ .unitPrice ]] \\
  [[ end ]]
  \bottomrule
\end{longtable}

\hfill Total due: [[ .total ]]

Output — invoice.pdf

Invoice

INV-2026-001

Bill to

Acme GmbH

94 Market St, San Francisco

Due date

2026-07-12

Description	Qty	Total
Backend development	10	1,200.00
API integration	4	560.00

Total due

EUR 1,760.00

Page 1 of 3 — headers repeat, totals never split

The problem

HTML-to-PDF breaks when pagination matters.

Print CSS works for simple one-page documents. It falls apart the moment real documents need structure.

Tables split badly across pages
Headers and footers are inconsistent
Totals land on wrong pages
Page breaks require hacks
Browser rendering differs across environments
Good multi-page layout takes too much time

How it works

Use data. Keep templates. Get PDFs.

JSON data + LaTeX template → PDF

Create a LaTeX template with placeholders: Design the document once. Merge fields, loops, and layout rules live in the template.
Send resources plus optional JSON data: One POST request with your LaTeX files and data for [[ ]] placeholders. No TeXLive, no Docker images.
Receive PDF bytes synchronously: Pixel-perfect pagination, repeated headers, page numbers, and totals where they belong.

Examples

Same API. Very different documents.

Each example shows the JSON payload, the template snippet, and the generated PDF.

Numbered sections, tables, charts, and an appendix with predictable page breaks.

JSON data

{
  "title": "Q2 Usage Report",
  "period": "2026-04 / 2026-06",
  "sections": [
    {
      "heading": "Summary",
      "metrics": [
        {
          "label": "Renders",
          "value": 48210
        },
        {
          "label": "p50 time",
          "value": "740ms"
        }
      ]
    },
    {
      "heading": "Appendix",
      "table": "usage_by_team"
    }
  ]
}

LaTeX template

[[ range .sections ]]
\section{[[ .heading ]]}
[[ if .metrics ]]
\begin{tabular}{l r}
  [[ range .metrics ]]
  [[ .label ]] & [[ .value ]] \\
  [[ end ]]
\end{tabular}
[[ end ]]
[[ end ]]
\appendix
\input{tables/usage_by_team}

Output — report.pdf

Q2 Usage Report

2026-04 / 2026-06

1. Summary

Renders	48,210
p50 render time	740ms

2. Renders by week

Page 1 of 6 — sections, tables, appendix

AI agents

Built for AI agents.

Texify gives agents a PDF tool they can call, inspect, and retry without guessing at browser print output.

Remote MCP server

Connect Claude Code, Cursor, or any streamable HTTP MCP client to https://api.texify.dev/mcp. No install.

Claude Code / Cursor

{
  "mcpServers": {
    "texify": {
      "url": "https://api.texify.dev/mcp"
    }
  }
}

Agent skill

Install the Texify skill from GitHub, then let the agent render, read compile errors, and fix LaTeX until the PDF compiles.

Install

opencode skill add https://github.com/texifydotdev

REST compile failures return 422 with a trimmed error snippet. MCP tool errors return the full LaTeX log verbatim, so agents can run the self-fix loop instead of asking humans to debug TeX.

View skill repo →

Developer API

Built for API workflows.

Render documents with a single request. REST compile errors include a trimmed snippet; MCP tool errors include the verbatim log.

Request

POST /render HTTP/1.1
Host: api.texify.dev
Content-Type: application/json

{
  "entrypoint": "main.tex",
  "resources": {
    "main.tex": "\\documentclass{article}\n\\begin{document}Hello Texify.\\end{document}"
  }
}

Response

HTTP/1.1 200 OK
Content-Type: application/pdf

%PDF-1.5
% binary PDF bytes...

Roadmap and trust

Typesetting first. Compliance next.

Texify is focused on reliable LaTeX rendering today. Production controls and compliance features are on the roadmap.

Available now

Raw LaTeX rendering
LaTeX templates
Trimmed REST compile errors
Remote MCP server
Template gallery

Coming next

PDF/A support, Ghostscript + veraPDF validated
Factur-X / ZUGFeRD e-invoicing
Hosted template-ID endpoint
Async jobs
Webhooks
Template versioning
Custom fonts
Image uploads

Exploring

AI compile-error explanation
HTML-to-LaTeX migration helper
Hosted template editor
Team accounts

Pricing

Pricing is being shaped with beta users.

Join the beta or request a paid pilot. We are using real document workflows to shape launch plans.

Beta access: Try LaTeX rendering and templates while the product takes shape.
Paid pilot: Get help turning one document flow into a working render path.
Launch planning: Share volume, compliance, and workflow needs for future plans.

Need this now? Request paid setup.

Send one template and sample JSON. We help convert it into a working Texify render flow.

Request paid pilot →

Join beta

Tell us about your PDF pain.

The more we know about your documents, the faster we can get your first template rendering.

Prefer to see output first? Download the sample PDF.

FAQ

Questions, answered.

Is this HTML-to-PDF?: No. Texify uses LaTeX-quality document rendering. HTML input may come later, but the current focus is structured data plus LaTeX templates.
Do I need to know LaTeX?: For the beta, yes. Later, templates and examples should reduce the LaTeX knowledge required to get started.
Can I use my own templates?: Yes. The beta supports user-provided LaTeX templates.
Can I send raw LaTeX?: Yes. POST raw LaTeX files as resources to /render and save the application/pdf response body.
Is this safe for private documents?: During beta, avoid sending highly sensitive production documents. Security and isolation are core roadmap items, but we will not claim enterprise security before it exists.
Why not use HTML-to-PDF?: HTML is great for simple documents. Texify targets documents where pagination, tables, math, and layout predictability matter.