FounderDuels

3 min read

A founder's guide to Pakistani payment rails in 2026

Raast, PayPak, Visa and Mastercard, JazzCash and Easypaisa, and possibly Apple Pay. What each one is actually good for, and what it costs you to support.

By the FounderDuels team ·

If you are building a product in Pakistan that takes money, you will end up supporting more rails than you expect. Here is the landscape as it stands in late 2026, with the trade-offs stated plainly.

Raast — the default you should start with

The State Bank's instant-payment rail. Account-to-account, real time, mostly initiated by QR at merchant checkout.

Scale: 742 million transactions worth Rs23.27 trillion in the third quarter of FY26. Person-to-merchant volume grew from 36.3 million to 55.9 million transactions quarter on quarter. More than 2.6 million merchants onboarded by March 2026.

Why you want it: it is effectively free at the point of transaction, settlement is instant, and it reaches the largest merchant base in the country. The State Bank put Rs3.5 billion behind QR adoption in FY2025-26 and daily P2M went from roughly 60,000 to about 1.1 million transactions in a year.

What it is bad at: recurring billing, card-on-file, pre-authorisation, and anything involving a foreign customer. It is a payment rail, not a card network, and it does not give you the chargeback and authorisation primitives that subscription products are built on.

Cards — Visa, Mastercard, PayPak

Visa and Mastercard are what Pakistani customers use for anything international, anything recurring, and most e-commerce. They carry interchange and scheme fees, in foreign currency.

PayPak is the domestic scheme, on more than a quarter of Pakistan's 53 million debit cards but carrying only about 6% of card usage. Cheaper for banks, domestic-only, and historically an ATM card in practice. 1LINK has said a PayPak credit card is intended for later in 2026, which would be the first domestic credit product.

What to support: if you have recurring revenue, you need cards. There is no way around it yet.

JazzCash and Easypaisa

Branchless banking with stored balances and agent networks. They reach customers the banking system does not, and they handle cash conversion, which nothing else on this list does.

Support them if your customers include people without a bank card, or if you need cash-in/cash-out. The international infrastructure layer is starting to treat them as first-class endpoints — EBANX announced in September 2026 that it is entering Pakistan with JazzCash, Easypaisa and Raast integrations planned.

Apple Pay — not yet

Apple Pay is not available in Pakistan. Pakistan is not on Apple's supported-countries list and Apple has made no announcement. There is strong circumstantial evidence of preparation — PayPak strings in the iOS 27 beta, reported bank testing — but no launch.

Plan for it as an addition, not a replacement. When it arrives it will almost certainly be card-only, probably international schemes first, and it will not run on Raast. If your checkout already supports cards, Apple Pay is a button on top of the rail you already have, not a new integration of its own.

The practical stack for a Pakistani startup

For most products, in this order:

  1. Raast QR for in-person and one-off payments. Cheapest, widest merchant reach.
  2. Cards if you have subscriptions, international customers, or need authorisation primitives.
  3. Wallets if your customer base extends beyond the banked urban segment.
  4. Apple Pay as a checkout button once it exists, layered on the card integration you already built.

The mistake to avoid

Building your business model on card economics because that is what the international payments literature assumes. In Pakistan, 3.4 billion of 3.7 billion retail transactions in a quarter went through digital channels, and the growth is overwhelmingly on a rail that charges close to nothing. A model that needs card interchange to work is a model built for a market Pakistan is deliberately steering away from.


Sources

Published by FounderDuels. Corrections: [email protected]

Something here wrong or out of date?

Tell us