For six years, “free UPI” helped India transform how it pays. The next phase begins on October 15, when a Merchant Discount Rate (MDR) of 0.4 per cent will apply to person-to-merchant transactions above ₹2,000.
Calling this the end of free UPI is incomplete. Consumers will still pay nothing. Person-to-person transfers remain free. Merchant payments up to ₹2,000 will attract no charge. Micro-merchants receiving up to ₹1 lakh monthly through UPI QR codes are exempt. For transactions of ₹75,000 and above, the fee is capped at ₹300. Essential categories such as railways, fuel and telecom will face a flat ₹5 charge above the threshold.
The reform does not place a toll on every scan. It replaces universal zero pricing with selective merchant pricing. About 96 per cent of merchant transactions will remain unaffected.
UPI has earned the right to make this transition. In August 2026, it processed 24.51 billion transactions worth ₹29.82 lakh crore. In FY 2025-26, UPI accounted for 84 per cent of India’s digital payment volume. Eighty-six per cent of merchant payments are below ₹500.
Yet scale does not make a payment system costless. Banks, payment apps and aggregators must fund servers, fraud monitoring, dispute resolution and customer support. The official FAQ cites an industry estimate of roughly ₹20,000 crore annually for maintaining UPI. The government’s incentive outlay for low-value merchant transactions was ₹1,500 crore. Subsidies aided adoption but are an uncertain foundation for infrastructure.
The economics supports a calibrated merchant charge. UPI is a two-sided network. Consumers benefit when many merchants accept it, while merchants benefit when many consumers use it. A 2025 Bank for International Settlements study concludes that setting consumer, merchant and interchange fees at zero is unsustainable without subsidies, cross-selling or another revenue source. Its model also finds that zero merchant fees can cause under-provision of fast-payment services.
Free at the point of use does not mean free to produce. International experience offers the same lesson. Brazil’s Pix keeps transfers free for individuals while permitting business charges. Merchant fees have typically been around 0.3 to 0.35 per cent. Thailand’s PromptPay keeps transfers up to 5,000 baht free and permits small fees beyond that level. The European Union requires instant-transfer charges to be no higher than ordinary-transfer charges.
The principle is clear: protect mass access while creating credible cost recovery.
Thailand’s threshold deserves attention, although the systems are not identical. At current exchange rates, 5,000 baht is approximately ₹14,400. That overstates the Indian equivalent because Thai incomes and prices are higher. Measured relative to per-capita GDP, the Thai threshold corresponds broadly to an Indian threshold of around ₹4,500. A ₹5,000 threshold would be defensible.
Raising the threshold would cushion medium-ticket merchants. It may not alter transaction counts because only about 4 per cent of merchant transactions fall outside the present exemptions. Its effect on fee revenue could nevertheless be considerable. Larger transactions represent more value, and MDR is value-based. NPCI should publish the value distribution above ₹2,000 before policymakers conclude that the threshold is optimal.
The larger problem is not only where the threshold is fixed, but how it operates. A ₹2,000 payment attracts no MDR. A ₹2,001 payment suddenly costs the merchant about ₹8 because the fee applies to the entire amount. This pricing cliff encourages bill splitting and substitution towards cash or other payment methods.
A marginal structure would be cleaner. The first ₹2,000, or perhaps ₹5,000, could remain free, with MDR applying only to the amount above the threshold. Graduated rates could then rise with transaction value. This would cushion medium-sized purchases while preserving cost recovery from commercial payments. It would also reduce the incentive to split one purchase into several transactions.
Implementation must address indirect pass-through. A prohibition on charging buyers does not prevent merchants from embedding MDR in general prices or offering cash discounts. Banks must also classify merchants accurately without penalising small sellers.
The revenue-sharing formula matters. Zero MDR favoured deep-pocketed platforms that could absorb losses and monetise adjacent services. A revenue stream can support smaller competitors and better fraud controls. It can also entrench incumbents if allocation is opaque. NPCI and the RBI should disclose transaction costs, revenue sharing, uptime, failure rates, fraud losses and grievance-resolution times. Part of the revenue should depend on measurable service improvements.
India should therefore resist two extremes. Permanently forcing every participant to provide UPI at zero price can weaken investment and competition. Treating UPI like a commercial card network can erode inclusion and trust.
This is not the end of free UPI for Indians. It is the end of an unlimited subsidy for larger merchant transactions. That transition is justified. However, a higher threshold, marginal charging and graduated rates would produce better economics than an abrupt pricing cliff. The fee is small, but the design and governance tests are not.
The author is a senior associate professor and director (accreditation) at Great Lakes Chennai
The opinions expressed in this article are those of the author and do not purport to reflect the opinions or views of THE WEEK.
