n8n Invoice Extraction from Gmail PDFs to Drive & Sheets

Automatically extract invoice details from Gmail PDF attachments—store PDFs in Google Drive and log everything in Google Sheets

This n8n workflow watches your Gmail inbox for incoming invoice PDFs, extracts the invoice text, uses Anthropic Claude to capture key fields as structured JSON, then saves the original PDF to Google Drive and appends a new record to your Google Sheets “請求書管理” sheet.

What this workflow does

• Polls Gmail every minute for emails matching has:attachment filename:pdf, then downloads the PDF attachment.
• Extracts text from the attached PDF to obtain the invoice content.
• Feeds extracted text to Anthropic Claude and requests a JSON response containing:
  • invoice_no
  • issuer
  • amount
  • due_date
• Builds a standardized filename using the invoice number and combines AI fields with email metadata (sender and received date).
• Uploads the original PDF to Google Drive in a specified folder (using OAuth2 credentials).
• Appends a row to Google Sheets in the sheet “請求書管理”, including: 差出人, 請求元, 受信日時, 支払期限, 請求金額, ファイルURL, 請求書番号.

Use cases

• Accounts payable teams who receive invoices via email can automatically centralize PDFs and maintain an always-up-to-date invoice register.
• SaaS operators can reduce manual data entry by converting invoice PDFs into structured records within Google Sheets.
• Automation engineers can use this n8n workflow as a reliable Gmail → PDF text extraction → Claude JSON parsing → Drive/Sheets pipeline.

Technical details

• n8n nodes/integrations: Gmail polling/trigger, PDF text extraction (extract from file), Anthropic Claude (n8nn8n-nodes-langchainanthropic), Code node, Google Drive, Google Sheets.
• Config requirements: Gmail OAuth2, Anthropic API credentials with JSON-only output format, Google Drive folder ID, and Google Sheets spreadsheet ID + sheet name with columns matching the mapped fields.