Which payment gateway to choose in Peru (and when you cannot choose)

Almost every comparison of Peruvian gateways opens with a fee table. It is the worst place to start: the percentages sit so close together that they decide nothing, they age within months, and many of the tables published today no longer match what the gateways themselves say. The decision gets made on other axes.

Let us start with the bias

We are a Solution Partner of several gateways. That means we have access to each one's technical team and gain nothing by pushing you toward any of them — but also that we are not a neutral observer, and you would rather know that before reading on.

We say it because none of the Peruvian comparisons ranking today declares whether it has a commercial relationship with the gateways it compares. Perhaps they have none. But when an article recommends one and does not say where that comes from, the reader has no way to calibrate it.

The fee tables you will find are already wrong

Not out of bad faith — out of expiry: gateways change rates and coverage, and articles do not. We checked against the official price sheets on the day we wrote this.

A concrete, verifiable example. Several of the best-ranked comparisons state that Culqi charges between 3.49 % and 4.20 % plus VAT, and that it only accepts Visa and Mastercard. Its public price sheet says otherwise: 3.44 % plus a fixed fee for domestic cards online, 3.99 % international, no onboarding cost and no monthly fee, same-day settlement, and it accepts Diners and American Express alongside Visa and Mastercard, plus Yape, Plin and BIM.

And a detail almost no comparison gets right: Culqi publishes its commissions as not subject to VAT. That is not a quirk — commissions charged by financial system entities and for collection services fall outside the tax. But the treatment depends on the provider and the type of operation, so the only way to compare properly is to read each price sheet and confirm it with your accountant, not to trust an article.

The opposite case is informative too: Niubiz does not publish an open price sheet. That is not an oversight, it is a commercial model — its rate is negotiated. A comparison that hands you an exact Niubiz figure should make you wonder where it came from.

That is why this article carries no fee table. The math that does matter —VAT, fixed fee, settlement delay and chargebacks— is worked through in what it really costs to charge online in Peru, where you will see why the advertised percentage is almost never what you end up paying.

The first question is not which, but whether you get to choose

If you are a public entity —a national university, a municipality, a government body— this comparison does not apply to you: you do not contract a commercial gateway, you collect through Banco de la Nación.

We crawled how the country's national universities collect today and did not find a single one with a gateway integrated into its own system. All of them route the payment to the bank. It is not a technology decision: public funds are handled through the national treasury system.

What you can decide is something else, and it is the part that matters. When the payment happens outside your systems, the proof arrives at the institution as a document, not as a record. And a document cannot be verified at scale: it gets looked at, stamped and filed. That is why receipts get forged — not because nobody checks, but because no office has anything to check against.

The fix is not changing banks: it is integrating that collection into your platform so any department validates the payment in real time. That is what we built for UNJFSC's central cashier, with validation against RENIEC. None of the comparisons competing for this query mentions this case, and in Peru it is a large share of the problem.

The six axes where the decision is actually made

If the percentage does not decide, what does? These six, roughly in the order they tend to weigh.

Settlement is working capital, not a detail

Between a gateway that settles same day and one that settles in three, there is a difference no fee comparison shows: how much of your own money sits permanently withheld.

If you bill S/ 300,000 a month, each day of delay is roughly S/ 10,000 floating outside your cash position at all times. Compare that with what half a point of commission saves you on the same sales —S/ 1,500— and the axis almost nobody looks at turns out to weigh several times more than the one everybody compares.

Amex and Diners: the fork almost nobody mentions

If part of your customers pay with American Express or Diners and your gateway does not accept them, that sale is not recovered by any rate negotiation. It is a yes or a no.

It matters most at high ticket, in travel and in corporate spend, where those brands have real presence. The check takes a minute and comes before any table: ask which brands it accepts today —not what an article said a year ago— and cross it with what your customers carry.

When more than one gateway makes sense

Almost never at the start. A second gateway doubles the reconciliation surface —two sources of truth, two webhook formats, two month-end closes— and that is paid in operations every month.

It is justified when there is a concrete, nameable reason: a card brand one of them does not take, a payment method only one supports, or volume large enough that a failover pays for itself. If you cannot name the reason in one sentence, you do not need it yet.

What to check before signing

In short

The question “which gateway is best” has no answer, and anyone who gives you one without asking about your operation is guessing. Start with whether you get to choose; if you do, decide on stage, settlement delay and card brands; and leave the commission for last, where the least is at stake. And if the charge also has to invoice itself or land in your ERP, that outweighs any tenth of a point.