Expenses

Extract from Receipt

Use AI-powered OCR to scan a receipt image or PDF and automatically fill in expense details.

Quick Start

Go to Dashboard -> Expenses and click Extract from Receipt.

Extract from Receipt button on Expenses page

PineBill uses AI to read your receipt and automatically fill in the expense form with the extracted details — title, amount, currency, date, category, vendor, and payment method.

Extract from Receipt uses AI credits. Each scan costs 2.5 credits. You can check your credit balance in the AI Credits settings.


How It Works

Extract from Receipt uses an AI vision model to read uploaded receipt images and PDFs. The model identifies expense-relevant information in the document and returns structured data that PineBill uses to pre-fill the expense form.

The AI extracts:

FieldDescription
TitleA short name derived from the receipt or vendor name
AmountThe final total paid or payable
CurrencyISO 4217 currency code detected from the document
DateTransaction date in the receipt
CategoryAutomatically classified into one of PineBill's expense categories
VendorMerchant or service provider name
Payment MethodCard, cash, bank transfer, or other method if visible
NotesAdditional context found in the document

The AI also extracts line items, tax details, and totals from the receipt for your review.


Supported File Types

FormatExtensions
ImagesPNG, JPG, JPEG, WebP, GIF
PDFPDF

Maximum file size: 5MB.


Step-by-Step Guide

1. Open Extract from Receipt

Go to Expenses and click Extract from Receipt. A dialog opens with a drag-and-drop upload area.

Extract from Receipt upload dialog

2. Upload Your Receipt

Drag a receipt file into the upload area, or click to browse your device. PineBill shows a preview of the uploaded file.

Receipt file preview before scanning

You can remove the file and choose a different one before scanning.

3. Scan the Receipt

Click Scan receipt to start the AI extraction. The dialog shows a scanning animation while the AI reads your document. This usually takes a few seconds.

Receipt scanning in progress

The credit cost is displayed before you confirm the scan. If you don't have enough credits, the scan button is disabled.

4. Review Extracted Details

After a successful scan, PineBill opens the Add Expense form pre-filled with the extracted data. You can review and adjust any field before saving.

The form shows an extraction review with:

  • Document type — what kind of document was detected (receipt, bill, purchase invoice, etc.)
  • Confidence scores — how confident the AI is about each extracted field
  • Warnings — any issues found during extraction, such as missing amounts or unclear text

5. Save the Expense

Review the pre-filled fields, make any corrections, and click Save Expense. The uploaded receipt is automatically attached to the expense.

Successful receipt scan result

Category Detection

The AI automatically classifies expenses into one of PineBill's categories based on the content of the receipt:

CategoryExamples
SoftwareSaaS subscriptions, hosting, domains, APIs, licenses
OfficeOffice supplies, stationery, coworking, rent
TravelFlights, hotels, taxis, trains, fuel, parking
MealsRestaurants, cafes, food, coffee, catering
MarketingAds, campaigns, design, branding, promotion
EquipmentComputers, phones, hardware, tools, furniture
ContractorsFreelancers, agencies, consultants
Taxes / FeesTaxes, government fees, bank fees, accounting fees
OtherAnything that doesn't fit the above categories

You can change the category after extraction if the AI classification doesn't match.


AI Credits

Each receipt scan costs 2.5 AI credits. Credits are deducted after the scan completes successfully. If the scan fails, no credits are charged.

Tips for Best Results

  • Use clear, well-lit photos or high-resolution scans
  • Make sure the entire receipt is visible in the image
  • Avoid blurry, folded, or partially cropped receipts
  • PDF receipts from digital purchases typically give the best results
  • If extraction fails, try a different photo angle or a higher quality scan

Common Questions


On this page