One API contract, one settlement file, thirteen ways for a shopper to pay. This page is the reference: what each method covers, what it does about refunds, and whether it can carry a subscription.
Methods are grouped by the rail the money actually travels on, not by how they look at checkout. That grouping is the one that predicts behaviour: everything on a card rail shares chargeback exposure and settlement timing, and everything on a bank rail does not.
Every method below is enabled with a toggle in the dashboard and reaches you through the same request shape and the same webhook. Adding one is a configuration change, not an integration project — which is the whole reason to run them from a single gateway rather than to collect providers per market.
Scheme rails. Highest reach, chargeback exposure, and the only family with a full dispute process.
Interchange++ pricing is published, so the scheme fee and the interchange arrive on the statement separately.
Network tokens are requested automatically for stored credentials, which lifts repeat-billing approval rates.
French cards are co-badged by law. Routing to the domestic scheme is usually cheaper and is on by default.
Debit only and never MOTO. Treat it as a card that always authenticates.
Card rails behind a device token. Authentication is delegated to the handset, so authorisation rates run high.
Biometric approval counts as strong authentication, so the payment is exempt from a separate 3-D Secure step.
Some tokens arrive unauthenticated. We step those up to 3-D Secure 2 rather than sending them through raw.
The shopper authorises inside their own banking app. No chargebacks, so no dispute exposure either.
Irrevocable once confirmed and outside the chargeback system. The most common first APM for a Dutch storefront.
Roughly one Belgian checkout in two. Skipping it in Belgium costs more conversion than any card optimisation wins back.
The code expires in two minutes. Keep the checkout page alive rather than redirecting away while it is entered.
Covers the Polish banks that sit outside BLIK. Usually enabled alongside it, not instead of it.
Austrian shoppers expect it next to card. The bank list is short, so the picker is not a friction point.
The cheapest rail we run and the one with no chargeback exposure. See the open banking product page for coverage.
You pull the money on a mandate. Cheapest per transaction, slowest to clear, reversible for eight weeks.
The shopper can reverse a collection for eight weeks with no reason given, and thirteen months if no mandate exists.
Enable by where the shopper is, not by where you are. The lead method is the one that should appear first at checkout; the rest are there so nobody is stuck.
| Market | Lead with | Then | Why |
|---|---|---|---|
| Netherlands | iDEAL | Cards, Apple Pay, Google Pay | iDEAL is the default expectation. A Dutch checkout without it loses more conversion than any card optimisation recovers. |
| Belgium | Bancontact | Cards, wallets, SEPA for subscriptions | Roughly one Belgian checkout in two. The app flow is fast and the method is irrevocable once confirmed. |
| Poland | BLIK | Przelewy24, cards, Google Pay | BLIK covers the phone-first majority; Przelewy24 catches the banks that sit outside it. Enable both. |
| Austria | EPS | Cards, wallets | Austrian shoppers expect EPS beside card. The bank list is short, so the picker adds almost no friction. |
| Germany | Cards and wallets | SEPA Direct Debit, pay by bank | German shoppers are direct-debit-literate. For subscriptions, SEPA usually beats card on both cost and churn. |
| France | Cards | Wallets, pay by bank | French cards are co-badged, so let domestic routing take Cartes Bancaires where it is cheaper. |
| Southern Europe | Cards | Wallets, pay by bank | Card-led markets with high wallet uptake on mobile. Wallet buttons matter more than an extra bank method. |
| Nordics | Cards and wallets | Pay by bank | High card penetration and high trust in bank redirects. Local schemes are on the roadmap, not shipped. |
A checkout showing eight methods to every shopper converts worse than one showing three chosen by billing country. Every cart connector we ship can filter by country, and the hosted checkout does it from the payer's billing address and IP without configuration.
Strong customer authentication is a PSD2 obligation, not a BazPay policy. What differs between gateways is how much of it a shopper actually experiences, because the regulation allows exemptions and applying them is a technical decision.
An exemption is a request, not an instruction: the issuer can still challenge, and some routinely do. The detail of which exemptions apply and what happens when one is refused is on fraud and SCA, and the API-level parameters are in the payments guide.
Refunds behave the same everywhere: full or partial, up to the captured total, initiated by you. Reversals are where the families genuinely differ, and that difference should influence which methods you promote.
| Family | Shopper can reverse? | Window | What it means for you |
|---|---|---|---|
| Cards | Yes, via chargeback | Up to 120 days, longer in some reason codes | You defend with evidence and can lose. Fees apply either way. |
| Wallets | Yes, same card rules | As cards | Device authentication usually shifts liability to the issuer. |
| Bank rails | No | Not applicable | Confirmed means final. A dissatisfied shopper must ask you directly. |
| Direct debit | Yes, without giving a reason | Eight weeks, or thirteen months with no valid mandate | Keep the mandate evidence. It is the whole defence. |
This is the practical case for leading with a bank method in the markets that have one: the transaction costs less, settles a day sooner and cannot be reversed against you. Cards remain necessary for reach and for anything sold cross-border — they simply should not be the only thing a Dutch or Polish shopper is offered.
Per-method pricing, including the interchange++ breakdown, is on the pricing page. For how these methods reach a specific storefront, go back to the integrations hub.
Cart plugins
Not sure this is the right route?
Send the platform, the version and the markets you sell into. An engineer answers with the plugin, the API path or an honest "not yet" — not a discovery call.