Invoice

With Mindee, you can automatically parse invoices and extract structured financial data using the pre-trained Invoice model available in the Catalog.

Watch our quick demo to see how you can easily set up your custom invoice model with Mindee:

Why use Mindee for Invoices?

Invoices can vary widely by supplier, format, and layout. The Invoice model is designed to handle these differences, so you get consistent and reliable data extraction without custom development.

Common use cases:

  • Accounts payable automation

  • Expense and cost tracking

  • Tax and compliance reporting

Two Ways to Start Building your Invoice Model

  • Open the Document Catalog in your dashboard and select “Invoice.”

  • The model comes pre-configured with the standard financial fields listed above.

  • You can start using it right away, or adjust the schema in Data Schema if you need additional fields.

  • Once selected, you can immediately test the model with your own invoices.

2. Build a Custom Invoice Model with the AI Agent

  • If your workflow requires extra fields (e.g. purchase order number, IBAN, payment terms), you can describe them directly to the AI Agent.

  • Optionally upload a sample invoice for context.

  • The Agent will generate a tailored schema that extends beyond the built-in Invoice parser.

You can use this invoice sample if you want to try and do a live test yourself:

a fake invoice from Turnpike Designs

What can be Extracted from Invoices?

Here is the list of fields that are already in our pre-trained invoice model:

Field
Type

Supplier Name

text

Supplier Phone Number

text

Customer Company Registration (number, type)

array(object)

Supplier Company Registration (number, type)

array(object)

Invoice Number

text

Date

date

Due Date

date

Payment Date

date

Total Amount

number

Total Net

number

Total Tax

number

Taxes (rate, base, amount)

array(object)

Line Items (description, quantity, unit_price, total_price, product_code, tax_amount, tax_rate, unit_measure)

array(object)

Document Type

classification

Locale (language, country, currency)

object

Customer Name

text

Customer Address (address, street_number, street_name, po_box, address_complement, city, postal_code, state/region, country)

array(object)

Shipping Address (same components)

array(object)

Billing Address (same components)

array(object)

Supplier Address (same components)

array(object)

PO Number

text

Reference Numbers

array(text)

Supplier Payment Details (IBAN, SWIFT/BIC, account_number, routing/bank identifier)

array(object)

Supplier Website

text

Supplier Email

text

Customer ID

text

Keep in mind that users can add their own fields or delete existing fields while reviewing the data schema.

Document format support

The Invoice model accepts PDFs and common image formats (JPG, PNG). It works reliably with scanned, photographed, and digital invoices.


Last updated

Was this helpful?