Docyield's receipt parser turns a photo, scan, or PDF of a receipt into structured data — merchant, date, line items, tax, tip, and total — that you can export as JSON, CSV, or Excel. Snap a picture of a crumpled till receipt and get back clean fields instead of a blurry image buried in your camera roll.
Receipts are among the messiest documents to read: thermal paper fades, text is tiny, layouts vary by point-of-sale system, and photos are taken in bad lighting at odd angles. Docyield is built for exactly this. It reads the receipt in context, separates the purchased items from the summary totals, and returns a record you can file, reimburse, or analyse.
What a receipt parser does
A receipt parser converts the printed slip into labelled data. It identifies the merchant at the top, the date and time of the purchase, each item with its price, and the financial summary — subtotal, tax, tip, and the final total — plus how the bill was paid. The output is consistent whether the receipt came from a restaurant, a supermarket, a taxi, or an online order printed to PDF.
The valuable distinction is between reading characters and understanding a purchase. OCR can transcribe the text on a receipt, but it will not tell you which line is the tax and which is an item that happened to be priced the same. Docyield makes those judgements and returns the tax in the tax field and the items in the items list, so the data is immediately usable.
Expenses, bookkeeping, and reimbursement
The most common use is expense capture. Employees photograph receipts and the structured data flows into an expense report or accounting system, so the merchant, date, and total are recorded without anyone retyping them. Itemised extraction goes further, letting you split a single receipt across categories — meals versus alcohol, or office supplies versus equipment — when policy requires it.
For bookkeepers, the same export drops cleanly into a spreadsheet or accounting tool. Because tax is pulled out as its own field, reclaiming VAT or sales tax becomes a matter of summing a column rather than reopening every image to find the number.
Who uses it
- Employees and finance teams capturing expense receipts for reimbursement.
- Bookkeepers recording purchases and reclaiming VAT or sales tax.
- Small businesses keeping a digital record of every receipt for tax time.
- Apps that let users scan receipts for budgeting, loyalty, or warranty tracking.
- Developers adding receipt capture to a product through the Docyield API.
How receipt OCR works
There are two steps behind the result. First, optical character recognition turns the pixels of your photo into text — recognising the shapes of letters and numbers even on a curled, low-contrast thermal slip. Second, that text is interpreted: Docyield works out which line is the merchant, which entries are purchased items, and which numbers make up the subtotal, tax, tip, and total.
The second step is what separates a receipt parser from raw OCR, and it is where the value sits. Plenty of tools will hand you the text of a receipt; far fewer will hand you a record that says the merchant is this, the total is that, and here are the items in between. Returning labelled fields means the data is ready for an expense report or a spreadsheet without a human deciding what each number means.
Itemised capture versus totals only
For simple expense tracking, the merchant, date, and total are often enough, and you can ignore the rest. But itemised capture unlocks more: splitting a supermarket receipt between personal and business items, separating reclaimable from non-reclaimable lines, or analysing exactly what was bought. Docyield returns both levels, so you can use as much detail as the task requires.
Itemisation also helps verification. When the line items, tax, and tip add up to the stated total, you have a strong signal the receipt was read correctly; when they do not, that is your cue to glance at the image. The detail is there to be used when it matters and quietly ignored when it does not.
Getting the best results from a photo
Receipts are physical and often imperfect, so a little care with the image pays off. A flat, well-lit photo taken straight on, with the whole receipt in frame, gives the parser the most to work with. Long receipts photographed in good light read far better than a crumpled slip shot in a dim restaurant.
When something on the receipt is genuinely unreadable — a faded line on old thermal paper, for instance — Docyield returns that field empty rather than guessing. For an expense record, a blank you can fill in beats a confidently wrong number that slips into your accounts unnoticed.
Keeping receipts for tax and records
Tax authorities and accountants want receipts kept, but a shoebox of fading paper is no one's idea of a record. Converting each receipt to data as you receive it builds a digital archive that stays legible long after the thermal ink has gone, with the merchant, date, tax, and total preserved as text you can search and total at year end.
Because tax is captured as its own field, the figures you need for a VAT or sales-tax return are already separated out rather than buried in an image. When it is time to file or to hand records to an accountant, you have a spreadsheet that adds up instead of a stack of photos someone has to read one by one.
Output formats and scaling up
Each parsed receipt can be exported as JSON for developers, or CSV and Excel for finance and bookkeeping. The structured result is the same across formats, so you can switch outputs without re-uploading the image.
The free tool processes one receipt at a time, which suits occasional captures. For an app that scans receipts continuously, or a finance team handling a monthly pile of them, the Docyield API and paid plans add batch processing and webhooks while returning the identical schema.
What the receipt parser extracts
Every receipt is returned against the same schema, so the fields line up whether the slip came from a cafe or a hardware store.
- Merchant
- The name of the store, restaurant, or vendor.
- Date
- The date of the purchase.
- Time
- The time of the purchase, when printed.
- Items
- Each purchased line with its description, quantity, unit price, and amount.
- Subtotal
- The total before tax and tip.
- Tax
- The tax or VAT amount.
- Tip
- The gratuity, where applicable.
- Total
- The final amount paid.
- Currency
- The currency code or symbol on the receipt.
- Payment method
- How the bill was paid — cash, card, and so on.
How to convert a receipt to JSON, CSV, or Excel
- 1Upload the receipt — drop a photo, scan, or PDF above, or click to choose a file.
- 2Wait a moment while Docyield reads the slip and separates items from totals.
- 3Review the merchant, total, and a couple of line items against the image.
- 4Choose your output — JSON for apps, CSV or Excel for expenses and bookkeeping.
- 5Download or copy the result and file it wherever it needs to go.
Frequently asked questions
Processing documents at scale?
Batch upload, an extraction API, and webhooks for 100+ documents a month.
