We went looking for petty cash software on 1 August 2026 and found something unusual: there is no category.
What comes back is expense-claim platforms (a different problem — reimbursing employees who spent their own money), general-purpose accounting software, ERP modules (a different scale), and at least one Windows desktop tool whose listing dates from 2015.
Two of those deserve naming, because they are free, well made, and worth your look before ours. HomeBank describes itself as "Free, easy, personal accounting, software for everyone". GnuCash is blunter — "GnuCash is powerful accounting software" — and it is not merely a household tool: its own features page publishes a set of "Small Business Accounting Features" covering customer and vendor tracking, invoicing, bill payment, and A/Receivable and A/Payable accounts (both read 5 September 2026). Neither is a petty-cash tool, and that is a statement about shape rather than quality: they keep double-entry books of accounts, where this problem is a float of notes held by a named custodian and counted against a tin on a Friday afternoon.
That absence is worth explaining before recommending anything, because the honest answer for many businesses is that they do not need software for this at all.
Why the category is empty
Petty cash sits in a gap. It is too small for the expense-management vendors, whose products exist to handle card spend, receipt capture and reimbursement workflows for dozens of employees. And it is too structured for personal finance tools, which model one person's accounts rather than a float held by a named custodian and reconciled against a physical tin.
So the market never formed, and the default tool became a spreadsheet — which is a perfectly reasonable outcome, right up until it is not.
What a spreadsheet genuinely does fine
Let us be clear, because most articles in this shape are not:
If you run one cash box, one person touches it, the float is small, and month-end reconciliation takes ten minutes, a spreadsheet is the correct tool and you should keep using it. It is free, everyone can read it, and it needs no server.
Below a certain size, software is overhead that exceeds the problem it solves. Nobody selling software says this. It is nevertheless true.
Where a spreadsheet stops being adequate
Four specific failures, and they are recognizable:
1. Multiple boxes, multiple custodians. The moment there are two tins — a site office and a workshop, a front desk and a kitchen — a shared spreadsheet stops being one person's record and becomes a coordination problem. Two people editing, one overwrite, and a month's reconciliation is gone.
2. Floating-point rounding. This sounds pedantic and it is not. Spreadsheets compute in floating point; a column of currency values summed repeatedly will eventually disagree with the tin by a cent or two. A running balance that is a cent out is worse than no running balance, because it destroys trust in the whole ledger — and you cannot tell whether the cent is a rounding artifact or a mistake.
3. No enforced audit trail. A spreadsheet cell can be changed and no record survives. For a cash system, that is the whole ballgame. Petty cash is the single most fraud-exposed process in most small organizations, not because people are dishonest but because it is physical money with weak controls.
4. The reconstruction problem. When the float does not match the tin, you need the sequence — what was disbursed, in what order, by whom, against what category. A spreadsheet gives you a list. It does not give you a balance you can trust at any point in time.
What a purpose-built ledger does differently
Cashora is a self-hosted petty-cash ledger — cash boxes with an imprest float, categorized disbursements, running balances and month-end reports. $59.00, one time (product page, read 2026-08-20), and one of the three products in the Finance Ops Suite.
Try the Cashora demo ↗Live, on sample data, no sign-up.
The two design decisions that matter are both in its own FAQ, and both address the failures above directly:
"How is the math kept exact?" — "All amounts are stored as integer cents and balances are summed in PHP, never in floating point — so the running balance and the month-end report always reconcile to the cent."
That is failure 2, solved at the storage layer rather than papered over with formatting. It is a small engineering decision and it is the difference between a ledger you trust and one you argue with.
"Can I run more than one cash box?" — "Yes — that's the core design. Each box has its own custodian, float, and currency, with its own running balance and month-end report."
That is failure 1 — and the imprest float model is the point. A float is topped back up to a fixed amount each period, so the arithmetic that must always hold is simple: cash in the tin + receipts = the float. A tool built around that invariant catches a discrepancy the moment it appears rather than at year-end.
And what it deliberately is not:
"Does Cashora move money?" — "No — it records cash movements only. No cards, no bank connections, no payment processing of any kind. It is a ledger, which is why it has zero payment/PCI surface."
For a tool handling money, "we do not touch the money" is a security feature, not a limitation. There is no card data to leak and no payment integration to compromise.
The honest limitations
It does not replace expense claims. If your actual problem is employees spending their own money and claiming it back, this is the wrong category — you want expense-claim software, and the self-hosted end of that market is thin in a different way.
It does not do receipt capture. No photographing receipts with a phone. You still keep the paper.
It does not connect to your accounting system. Month-end reports come out; you enter the totals wherever your books live.
It is still a manual record. Software cannot make somebody write down a disbursement they did not write down. No petty-cash system solves the human failure at the tin — it only makes the record that is kept exact, multi-user, and reconcilable.
The honest costs of self-hosting
Backups are the ledger. Losing the database loses the audit trail, which is the entire reason to move off the spreadsheet. Test a restore before you trust it.
No vendor on call. For a petty-cash ledger this risk is genuinely low — if it is down for a day, the tin still works and you write on paper — but it is not zero.
Hosting. "SQLite for a zero-config trial, MySQL/MariaDB for production on ordinary shared hosting" per the product page. If you already run a website, the marginal cost is often nothing. For a concrete number from a named provider: Hetzner's own pricing — machine-read on 10 August 2026 from the price API behind its web-hosting and cloud pricing pages — lists web-hosting plans from €1.60 to €16.72 a month and entry cloud servers from €3.99 a month (prices as returned for their German datacenters; confirm current pricing and VAT treatment at checkout). Any comparable host works — the point is that the "own" side of this comparison costs single-digit euros a month.
Who should actually move off the spreadsheet
Yes, if any of these is true: two or more cash boxes · more than one custodian · a float large enough that a discrepancy matters · an auditor, funder or franchisor who asks to see the trail · you have already had a reconciliation that could not be explained.
That last one is the honest trigger. Most organizations move when something has already gone wrong once.
No, if: one box, one person, a small float, and month-end takes ten minutes. Keep the spreadsheet. We would rather say that than sell you a $59 license you do not need.
The short version
There is no petty cash software market because the problem sits between two categories that both ignore it. The default is a spreadsheet, and for a single tin with a single custodian the spreadsheet is genuinely the right answer.
The moment there are two boxes, two custodians, or an auditor, the spreadsheet's three real weaknesses — floating-point drift, silent edits, and no reconstructable sequence — stop being theoretical.
A purpose-built ledger that stores integer cents, models an imprest float per box, and touches no money at all costs $59 once. That is the entire pitch, and it does not apply to everyone.
Next steps
- Open the live Cashora demo — no sign-up, already signed in, resets itself
- Cashora product page and license options
- Self-hosted office and finance tools
This guide quotes no competitor prices because, unusually, there are no directly comparable products to quote. Where a category is empty, we say so rather than manufacturing a comparison.