Guide · n8n

Generate PDF documents from n8n

Add a document step to any n8n workflow with the n8n-nodes-docmill community node: JSON in, a finished PDF, DOCX, or XLSX out, ready to email, upload, or archive — no HTML wrangling inside your workflow.

What you need

• An n8n instance (n8n 1.0 or later — the node has no external runtime dependencies).
• A Docmill API key. The free tier includes 50 documents a month (watermarked), no credit card required.

Step 1 — Install the community node

1. Open Settings → Community Nodes

   On a self-hosted n8n instance, go to Settings → Community Nodes → Install a community node.

2. Enter the package name

   Type n8n-nodes-docmill, tick the box acknowledging the risks of installing community nodes, and click Install.

3. Find the node

   After the install finishes, search for Docmill in the nodes panel. On n8n Cloud, community nodes become searchable once they are approved by n8n — if you don't see it there yet, use a self-hosted instance.

Step 2 — Create the Docmill credential

Add a Docmill node to a workflow and click Create new credential (or go to Credentials → Add credential → Docmill API). Two fields:

n8n verifies the credential against GET /v1/usage when you save it, so a bad key fails immediately rather than mid-workflow.

Step 3 — The node's operations

You normally run Generate Template (AI) once while building the workflow, note the template ID it returns, then use Generate Document with that ID on every execution. Every template is plain HTML underneath, so you can inspect and tweak it.

A worked example: webhook → invoice PDF → email

Three nodes: a Webhook trigger receives order data, Docmill turns it into an invoice PDF, and Gmail (or any email/storage node) sends it.

1. Webhook trigger

   Create a Webhook node (method POST). Suppose callers send {"client_name": "Northwind Traders", "item": "Design sprint", "price": 4800}.

2. Docmill — Generate Document

   Operation Generate Document, Template ID starter:invoice, Output Format PDF. Switch Data (JSON) to expression mode and map the incoming fields:

   {
     "client": { "name": "{{ $json.body.client_name }}" },
     "items": [
       { "description": "{{ $json.body.item }}",
         "quantity": 1, "unitPrice": {{ $json.body.price }} }
     ],
     "total": {{ $json.body.price }}
   }

   The node outputs { success: true, bytes, format } as JSON and the file itself as the binary property data (filename document.pdf).

3. Send it

   In a Gmail / Send Email node, add an attachment and point it at the binary property data. The same property works for Google Drive, Dropbox, S3, Slack, and any other node that accepts binary input. If you chain several Docmill nodes, give each one a distinct Put Output File in Field name.

What the node does under the hood

One authenticated POST. If you'd rather call the API from an HTTP Request node (or anywhere else), this is the entire integration:

# one POST, one invoice
curl https://api.docmill.io/v1/documents \
  -H "Authorization: Bearer dk_live_…" \
  -H "Content-Type: application/json" \
  -d '{
    "template_id": "starter:invoice",
    "data": { "client": {"name": "Northwind Traders"},
              "items": [{"description": "Design sprint",
                         "quantity": 1, "unitPrice": 4800}],
              "total": 4800 },
    "output": "pdf"
  }' --output invoice.pdf
# → %PDF-1.7 …

The n8n node calls the same endpoint with ?encoding=base64 and decodes the response into the binary property for you. Swap "output": "pdf" for "docx" or "xlsx" to get Word or Excel files from the same template — in the node that's just the Output Format dropdown.

Sync vs. async

The node renders synchronously: each item waits for its file. That's the right default for webhook-triggered flows and small batches.

For bulk runs (say, hundreds of invoices from a Google Sheets read), the API also has an async mode: add ?async=1 to the documents call and you get a job ID back immediately; poll the job until it's done. In n8n you'd use an HTTP Request node plus a Wait node:

# enqueue
curl "https://api.docmill.io/v1/documents?async=1" \
  -H "Authorization: Bearer dk_live_…" \
  -H "Content-Type: application/json" \
  -d '{ "template_id": "starter:invoice", "data": { … }, "output": "pdf" }'
# → { "job_id": "job_a1b2c3…", "status": "queued" }

# poll until status is "done", then decode data_base64
curl https://api.docmill.io/v1/jobs/job_a1b2c3… \
  -H "Authorization: Bearer dk_live_…"
# → { "job_id": "job_a1b2c3…", "status": "done",
#     "content_type": "application/pdf", "data_base64": "JVBERi0…" }

Troubleshooting