docforge

DocForge — document data extraction for n8n. Turn any PDF, image, or Office file into structured JSON with confidence scores.

Package Information

Released: 1 mo ago
Downloads: 0 weekly / 0 monthly
Latest Version: 0.1.1

Documentation

n8n-nodes-docforge

DocForge community node for n8n — turn any document into structured JSON with per-field confidence scores.

Send a PDF, image, or Office file; get back structured data (invoice totals, line items, dates, parties...) plus a confidence score and arithmetic consistency checks (subtotal + tax = total, etc.).

Operations

Operation What it does
Extract Document Upload a file and extract data. Waits for the result by default (one call, up to 40s). Auto-detect, template, spreadsheet layout, JSON Schema, or plain-English instructions.
Get Task Status Poll a task returned as 202 (used when wait times out).
List Templates The 95 built-in extraction profiles (invoice, receipt, bank_statement, w2_tax, shipping_label, ...).
Get Account Info Plan, usage, quota.

Setup

  1. In n8n: Settings → Community nodes → Install → enter n8n-nodes-docforge
  2. Add credentials of type DocForge API:
  3. Feed a file in (e.g. Read Binary File from Disk), connect a DocForge node, run.

Example

Read Binary File → DocForge (Extract, Template: invoice) → whatever comes next.

Output shape:

{
  "status": "success",
  "waited": true,
  "result": { "invoice_number": "INV-000004", "total": "343.35", "line_items": [ ... ] },
  "confidence": { "score": 90, "level": "high", "consistency": { "all_passed": true } }
}

Use confidence.score >= 80 && confidence.consistency.all_passed !== false to auto-accept; route everything else to human review.

Limits & notes

  • Files up to 25 MB (5 MB on the free plan): PDF, JPG, PNG, TIFF, BMP, DOCX, PPTX, HTML, MD, TXT
  • With Wait for Result on, one document = one API request (quota-friendly)
  • Deterministic: same document + same mode = same result
  • Docs: docforge.biz/docs — Support: hello@docforge.biz

License

MIT

Discussion