Payment Gateway Fees in Nepal Compared

Every Nepali merchant who accepts digital payments pays a fee, and almost none of them know their real rate. Not the rate on the pricing page - the effective rate, the one that shows up after rounding, category rules, promotional expiries, and the occasional error. This article compares fees across eSewa, Khalti, and Fonepay, and then shows you the more important skill: auditing what you were actually charged against what you should have been.

MDR: the number behind the fee

The fee you pay is a merchant discount rate, or MDR - a percentage of each transaction that the gateway keeps for moving the money. In Nepal, MDR is shaped by a few forces:

• Your merchant category. A utility biller, a retailer, and a large e-commerce platform are not charged the same. Category drives the base rate.
• Your negotiated terms. Volume gets you leverage. A merchant doing serious daily numbers rarely pays list price.
• Regulatory context. Digital payment pricing in Nepal operates under the eye of Nepal Rastra Bank, which periodically weighs in on interchange and merchant charges. Rates are not purely a free-for-all.

Because of all this, there is no single honest "eSewa charges X%" number. There is your rate, on your account, for your category - and it can differ from your neighbour's shop selling the same goods.

How the three gateways compare in practice

Rather than quote rates that will be stale by the time you read this, here is how the three differ in ways that actually affect reconciliation:

• eSewa. Fees are deducted before settlement to your bank, and the settlement arrives as a net lump. The wallet report is your source of truth for gross and fee. The challenge is that per-transaction fee visibility is thinner, so you often infer the fee from the gap between gross and net. See how to read an eSewa settlement report for the decomposition. Developer details live at developer.esewa.com.np.
• Khalti. Fees are also deducted before payout, but Khalti's documentation exposes per-transaction fee data on lookup. This makes Khalti the easiest of the three to audit at the transaction level - you get the charged fee directly rather than backing it out.
• Fonepay. Fonepay is the inter-bank QR network, so the economics involve the banks on both sides of the QR. For a merchant, the fee experience depends heavily on your acquiring bank's arrangement, which means your effective rate is a bank conversation as much as a Fonepay one.

The takeaway is not "gateway A is cheapest." It is that the three expose their fees with different levels of transparency, and that changes how hard you have to work to verify them. We compare the broader trade-offs in eSewa vs Khalti vs Fonepay for merchants.

Why your effective fee drifts

The rate on your agreement is a starting point, not a constant. Real effective fees move because of:

• Rounding. A fee computed to two decimals and then rounded, transaction by transaction, accumulates a small bias over thousands of payments. Individually invisible, collectively real.
• Category reclassification. If the gateway moves your merchant category, your base rate moves with it - sometimes without a loud announcement.
• Promotional rates expiring. An introductory rate that reverts to standard is one of the most common causes of a "sudden" fee increase that was actually scheduled all along.
• Tiered volume thresholds. If your rate steps down after a monthly volume, the effective rate within a month is not flat - it depends on where in the month a transaction fell.

Add these up and your effective monthly fee rate is a moving target. Assuming a flat rate is how merchants quietly overpay.

Auditing computed-versus-charged

Here is the discipline that turns fees from a mystery into a controlled cost. For every settlement period:

• Compute the expected fee on each transaction from your agreed rate and category rules. This is the fee you should have paid.
• Extract the charged fee - directly where the gateway exposes it (Khalti), or by decomposing gross-minus-net where it does not (eSewa).
• Compare, transaction by transaction. Small consistent gaps point to a rounding rule. A step change on a date points to a rate change. A one-off outlier points to an error worth disputing.
• Roll up the effective rate. Total charged fees divided by total gross gives your real blended rate for the period. Track it month over month and drift becomes visible.

This is also where tax hygiene enters. Fees and the VAT treatment of digital payments both affect your books - VAT on digital payments in Nepal and the Inland Revenue Department guidance are worth reading alongside your fee audit, because a fee is a deductible cost only if you can evidence it.

What a clean fee audit gives you

When you audit computed-versus-charged every period, three things happen. You catch overcharges while they are still disputable rather than months later. You know your true cost of accepting digital payments, which is a real input to your pricing. And you can prove your fee expense to an auditor with source rows, not an assumed percentage.

The reason few merchants do this is not that it is conceptually hard - it is that doing it by hand across three gateways, every settlement, every month, is relentless. This is precisely the drudgery RakamHQ removes: it computes the expected fee on every transaction, compares it to what each gateway actually charged, and flags the drift - so your effective rate is a number you know rather than a number you hope.