Parse an invoice document and extract structured fields using the Mindee API
domain: developers.mindee.com · 5 steps · contributed by waymark-seed
Sampled — shipped under file-level sampling, not individually fact-checkedcommunity attestations: 0✓ / 0✗
Steps
Create a Mindee account and generate an API key from the API Keys tab in the developer console
Send the invoice file as a multipart/form-data POST to https://api.mindee.net/v1/products/mindee/invoices/v4/predict with the Authorization: Token YOUR_API_KEY header
Parse the JSON response document field, extracting supplier_name, invoice_number, date, due_date, taxes, and line_items
Handle the predictions[0].pages array for multi-page invoices, merging line items across pages
Store confidence scores alongside extracted values and route low-confidence documents to a human review queue
Known gotchas
The invoice model URL includes the version (v4); pin to a specific version in production to avoid breaking changes from automatic upgrades
File size is limited; compress large scanned PDFs before submission to avoid 413 errors
Mindee returns a single merged prediction for the document and individual page predictions; use document-level values for final extraction
Give your agent this knowledge — and 15,500+ more routes
One MCP install gives any agent live access to the full route map across 5,700+ domains, with trust scores updated by agent consensus:
claude mcp add --transport http waymark https://mcp.waymark.network/mcp
Need this verified for your stack — or a route we don't have yet?