Docweave vs PDFShift
PDFShift is a mature HTML-to-PDF API that many teams already rely on. Docweave covers the same core job — HTML in, a real PDF out — and adds an MCP-native generate_pdf tool, per-document pricing, and template + JSON rendering. If you are building for AI agents or want to store a layout and bind data into it, that is where Docweave differs.
Where Docweave differs
| Docweave | PDFShift | |
|---|---|---|
| MCP-native tool | generate_pdf, open source (npx @docweave/mcp) | REST API |
| Pricing model | Per document | See provider pricing |
| Sources | HTML, URL, template + JSON | HTML, URL |
| Rendering | Chromium | HTML-to-PDF engine |
Switching from PDFShift
The request shape is familiar — POST a source, get a PDF back:
curl https://docweave.dev/api/v1/pdf \
-H "Authorization: Bearer dw_live_your_key" \
-H "Content-Type: application/json" \
-d '{
"source": { "type": "html", "html": "<h1>Invoice #1024</h1>" },
"options": { "format": "A4", "margin": "20mm" }
}' \
--output invoice.pdfFAQ
Is Docweave a drop-in PDFShift alternative?
For HTML-to-PDF, the shape is familiar: you POST a source and get a PDF back. Docweave accepts raw HTML, a URL, or a template + JSON, renders with Chromium, and returns the file. The main additions are a first-class MCP server and per-document pricing.
What does Docweave offer that a standard HTML-to-PDF API does not?
An open-source, MCP-native generate_pdf tool (npx @docweave/mcp) that any MCP client — Claude, Cursor, and others — can call directly, plus template + JSON rendering so you can store a layout once and bind data into it.
How does pricing compare?
Many legacy PDF APIs bill per page or per conversion credit. Docweave bills per document, so a long report and a one-page receipt are priced the same way. Check each provider's current pricing page for exact numbers.
Can I use Docweave from an AI agent?
Yes. Add the Docweave MCP server and your agent gets a generate_pdf tool — one tool call instead of HTTP plumbing, with built-in idempotency so it never double-generates the same document.
Try the MCP-native PDF API.
Get an API key