Skip to main content

​
What is a field?

A field represents one piece of information inside a schema. Examples:
  • Invoice number
  • Issue date
  • Total amount
  • Vendor name
Fields are the most important part of data quality. Clear fields lead to better results and less review.

​
Field properties

Every field in anyformat has at least three properties:
A human-readable label that still makes sense outside anyformat.Examples: invoice_number, issue_date, total_amount
Tells anyformat what kind of value to expect (text, date, number, etc.).This improves consistency, validation, and output quality.
Plain-English guidance for how to extract the value.Example: “Extract the final total including taxes. Ignore subtotals.”

​
Field types overview

Each field has a type. Field types help anyformat understand how to interpret, validate, and output values.
Field typeWhen to useExample
TextFree-form informationCompany name, address
DateCalendar datesInvoice date
Date & timePrecise timestampsTransaction time
Decimal numberNumbers with decimalsTotal amount, tax
Integer numberWhole numbers onlyQuantity, item count
Yes / No (Boolean)True / falseIs paid?
SelectOne value from a listStatus: Paid/Unpaid
MultiselectMultiple values from a listTags/categories
Object (Subtable)Repeating structured itemsLine items in an invoice

​
Field type best practices

Best for: Values that vary widely (company names, addresses)Watch out for: Don’t use if values are from a known list — use Select instead
Best for: Calendar dates when time isn’t relevantWatch out for: If time appears in the document, prefer Date & Time
Best for: Precise timestamps when documents include timeWatch out for: Ambiguity with timezones and format variations
Best for: Money, percentages, measurementsWatch out for: Avoid if values should be whole numbers
Best for: Counts, quantities, item numbersWatch out for: Not suitable for currency
Best for: Clear true/false questions (Is paid? Is signed?)Watch out for: Avoid vague cues like “maybe” or “often” — make instructions explicit

​
Complex field types

​
Select fields

Select fields are used when a value must be one of several predefined options. Example:
  • Status: Paid, Unpaid, Overdue
Why use Select?
  • Enforces consistency
  • Avoids spelling variations
  • Makes analytics and filtering reliable
Best practices:
  • Keep options short and unambiguous
  • Avoid overlapping meanings
  • Prefer Select over Text when values repeat

​
Multiselect fields

Multiselect fields allow multiple options to be selected. Example:
  • Document categories: Invoice, Receipt, Contract
Use Multiselect when:
  • More than one value can apply
  • Order does not matter
Best practices:
  • Limit the number of options
  • Avoid using Multiselect for free-form tagging
  • If values are mutually exclusive, use Select instead

​
Object (Subtable)

Object (Subtable) is for repeating groups of fields — basically a table of rows. Think:
  • An invoice has multiple line items
  • Each line item has the same structure (description, quantity, unit price, total)
Example schema pattern: 
line_items (Object/Subtable)
  ├── description (Text)
  ├── quantity (Integer number)
  ├── unit_price (Decimal number)
  └── line_total (Decimal number)
Use Object/Subtable when:
  • You see the same set of fields repeating
  • The document contains a list/table where each row is one “item”
  • You need structured output per row (not a blob of text)
Best practices:
  • Keep subtable fields minimal at first (3-5 columns)
  • Use clear row-level instructions: “Extract one row per item. Ignore headers and totals.”
  • Add shared fields like currency at the top-level (not inside each row)
  • Start with the most reliable columns first (description + amount), then expand
Common mistakes:
  • Using Object/Subtable for a single nested object (like a “vendor address” group). If it’s not repeating, it usually shouldn’t be a subtable.
  • Making the subtable too wide too early (10+ columns increases ambiguity)

​
What’s next?

Instructions

Learn how to write effective extraction instructions

AI Suggestions

Use AI to help build your schemas faster