Stripe vs Adyen vs Checkout.com: A 2026 PSP Pricing Teardown

The PSP RFP that arrives on most operator desks compares three columns: Stripe at 2.9% + 30¢, Adyen at "interchange++ plus 0.60%," and Checkout.com at "contact sales." The finance team picks the lowest number. Six months later the effective cost is 80bps higher than projected, chargeback economics turn out to be unfavorable, and nobody remembered to model the FX markup on cross-border volume.

Comparing PSPs by headline rate is one of the most expensive mistakes in payments procurement. The actual cost equation is landed effective cost per successful dollar — which means authorization-rate uplift, scheme fees pass-through, dispute economics, FX stacking, and whether you clear the volume bar for a contract that matches the marketing page. This piece tears down what each of the three actually charges in 2026, what model it uses, and where each one wins.

A note on where these rates come from. PSP pricing pages localise by IP address and cookie, not by URL path — stripe.com/pricing and stripe.com/en-us/pricing both served Singapore-denominated figures during this refresh, and only stripe.com/en-gb/pricing reliably showed UK numbers. Every rate below is labeled with the country and card type it applies to; if you check a vendor's page yourself, confirm which country's rate card you're actually looking at before quoting it.

The Three-Layer PSP Cost Stack

Before any comparison, the MDR on a card transaction is three layers stacked into one number:

1. Interchange — paid to the issuing bank, set by Visa/Mastercard, regulated in the EU and UK, uncapped in the US.
2. Scheme fees — paid to Visa/Mastercard themselves; 30+ line items, up 20–30% since 2019.
3. Acquirer / PSP margin — the only layer the PSP actually sets.

A blended PSP like Stripe absorbs all three into one published percentage. An IC++ PSP like Adyen passes interchange and scheme fees through at cost and charges its margin separately. The two models look comparable on a quote sheet but behave very differently as merchant volume, ticket size, and card mix shift.

This matters because card mix dictates which model wins. Regulated consumer debit in the EU caps interchange at 0.20%, and regulated consumer credit at 0.30%, under Regulation (EU) 2015/751 — on that traffic, IC++ is dramatically cheaper than blended. Commercial and corporate cards sit outside this entirely: they're excluded from the regulation's Chapter II caps, so interchange on B2B card volume is uncapped and routinely runs well above the regulated rate. A merchant with a heavy corporate-card mix is modeling a different cost curve than one with consumer debit, and the EU cap doesn't help them. Premium credit cards in the US push interchange above 2% — on that traffic, blended rates with included scheme fees can be cheaper than IC++ once the acquirer margin is added. The "right" PSP depends on a card mix the operator is rarely modeling carefully.

Stripe: The Published Rate Card

Pricing. Stripe publishes per-region rate cards; the figures below are the US and UK rate cards specifically (see the note above on geo-localisation). In the United States, card-not-present runs 2.9% + $0.30; manual entry adds 0.5%. In the United Kingdom, the structure tiers by card type: 1.5% + 20p for standard UK domestic, 2.8% + 20p premium UK, 2.5% + 20p EEA, 3.15% + 20p international. Card-present (Terminal) is 2.7% + $0.05 US domestic.

Stacking surcharges. A US-based merchant taking a euro card pays 2.9% + 30¢ + 1.5% international + 1% currency conversion. That is an effective ~5.4% on cross-border traffic, a number rarely modeled in procurement decks built off the 2.9% headline.

Disputes, fraud, and payouts. Stripe charges two separate dispute fees, not one: a dispute-received fee ($15 US / £20 UK) on every dispute regardless of outcome, plus a dispute-countered fee ($15 US / £20 UK) only if the merchant responds — refunded on a win, forfeited on a loss. A contested-and-lost dispute therefore costs double the headline figure some procurement decks model. Stripe's Smart Disputes service separately charges 30% of the disputed amount, but only on a win. Radar fraud screening runs $0.05 per screened transaction pay-as-you-go in the US (£0.04 UK); the paid Standard/Plus/Pro tiers are also sold as monthly subscriptions starting at $10 / $14 / $20 per month (US) and £8 / £12 / £17 per month (UK) — their per-transaction pricing loads client-side and isn't published in a country-stable way, so it's left out here rather than guessed at; Radar Lite is bundled free with standard pricing. 3DS2 is included free on standard pricing; custom pricing accounts pay $0.03 per 3D Secure attempt (US) / £0.03 (UK). Payouts are free on standard rolling/weekly/monthly schedules; instant payouts cost 1.5% of volume with a $0.50 minimum in the US and 1% of volume with a 40p minimum in the UK — a cost line the Adyen comparison below shouldn't get to skip either.

Pricing model. Blended publicly. Custom contracts shift to interchange-plus / IC++ once volume and negotiating leverage make a custom deal worth pursuing — Stripe doesn't publish where that line sits. Below it, the published page is what you pay.

Sweet spot. SaaS, subscriptions, marketplaces (Connect), platforms launching fast, embedded finance (Treasury, Capital, Issuing). Best-in-class developer experience and the fastest ramp from zero to live on the market. Stripe was named a Leader in The Forrester Wave™: Merchant Payment Providers, Q1 2026.

2026 differentiators. Stripe Authorization Boost — AI-driven retries, Data-Only authentication, PINless debit retries — claims +3.8% acceptance and −3.3% processing cost. Network token and card account updater coverage expanded across Visa, Mastercard, and Amex. Sessions 2026 launched Checkout Studio, agentic-commerce APIs, and stablecoin rails.

Adyen: Indicative IC++

Pricing. Adyen publishes an "indicative" rate card: a $0.13 base fee per transaction plus IC++ with an acquirer markup that starts around 0.60% and scales down with volume. That $0.13 fixed fee is quoted in USD uniformly across Adyen's own pricing table — even on EUR, GBP, SEK, and NOK-denominated rows — unlike Adyen's other fixed fees (SEPA Direct Debit, Klarna) which do show in local currency; worth knowing before modeling exact landed cost in a non-USD market. Alternative payment methods are priced separately — Klarna at $0.13 + 0.99–4.99% region-dependent, Alipay at $0.13 + 3%, Apple Pay at $0.13 + card fees passed through. Multi-currency settlement is offered to avoid FX where possible.

What's hidden. Disputes, 3DS / RevenueProtect pricing, payout/withdrawal fees, volume tier breakpoints, and business eligibility are all contact-sales. On the minimum invoice, Adyen's own pricing page states only: "We do have a minimum invoice depending on industry or business model. Please speak to a member of our sales team" — no figure is published, and no direct-contract volume threshold is disclosed anywhere on Adyen's site.

Pricing model. Pure IC++ with full pass-through transparency. Quote-based at the contract layer.

Sweet spot. Enterprise omnichannel — unified online + in-store + mobile on a single platform with one contract and one reconciliation feed. Single global acquirer in 30+ markets. Wins for retailers, airlines, travel, and marketplaces with physical + digital footprints (Uber, McDonald's, eBay, Patagonia). FY 2025 processed €1,394.3B in volume (+21% excluding one large customer departure) and €2,364.2M net revenue (+18%). Adyen's BFCM 2025 weekend processed $43B in volume with a 199,000-transactions-per-minute peak at 99.9999% uptime.

2026 differentiators. Adyen Uplift — AI suite generally available since 2025 — claims +6% authorization and an 86% reduction in manual fraud rules; piloted by Patagonia, Indeed, Nord Security, Fubo. Adyen has issued 2B+ active network tokens platform-wide, reporting an average 3% authorization-rate uplift for businesses using network tokenization; a machine-learning layer chooses token vs. PAN per transaction, though Adyen doesn't publish a separate incremental-auth figure for that specific feature. Adyen for Platforms competes directly with Stripe Connect on embedded finance.

Checkout.com: No Published Rate Card

Pricing. Public pricing page positions "transparent IC++" and "simple interchange++ fees based on business profile and risk" — and discloses no percentage, fixed fee, or range of any kind, for any product line. That absence is itself the finding: a PSP marketing "transparent" pricing while publishing zero numbers anywhere is a materially different posture from Stripe's published rate card or even Adyen's indicative one. Checkout.com publicly focuses on merchants at meaningful annual volume; SMB applicants are typically referred elsewhere.

Pricing model. Interchange++ with flat acquirer markup, fully quote-based. Effectively enterprise-only.

Sweet spot. High-volume card-not-present, global digital-native enterprises, marketplaces, crypto, gaming, travel, streaming, BNPL providers. Strong on Middle East and APAC where Stripe and Adyen have thinner local acquiring. Direct acquiring in 50+ currencies. Processed $300B+ in 2025 with 63 merchants doing $1B+/year each, including confirmed names Uber, eBay, Spotify, Temu, Pinterest, HelloFresh, ASOS, and Vinted. Returned to full-year EBITDA profitability with 10%+ margin. Valuation reset to $12B in September 2025 from the 2022 $40B peak.

2026 differentiators. Intelligent Acceptance has unlocked a claimed $10B+ in merchant revenue since its 2023 launch, with 60M real-time optimizations per day, an average +3.8% acceptance lift, and up to +9.5pp in best cases (Reach reported ~10% auth uplift). Flow Remember Me launched in October 2025 for one-click global checkout. Issuing reached a $5B run-rate by Q4 2025, expanding into the US and UAE through 2026.

The Pricing-Comparison Trap: Seven Mistakes Operators Make

1. Comparing headline % without scheme-fee pass-through. Adyen's "0.60% + IC++" sounds more expensive than Stripe's "2.9% + 30¢" until you realize Stripe's blended rate hides interchange + scheme markup that on regulated EU debit can be 5x what a true IC++ contract would cost. The comparison only holds at matching card mix.

2. Ignoring FX markup. Stripe stacks +1.5% international + 1% (US) or 2% (UK/SG) currency conversion. On a euro-denominated card hitting a US merchant account, the effective rate is ~5.4%. Most operators don't see this in the procurement deck because nobody added the international + FX line items to the 2.9% headline.

3. Missing the minimum monthly invoice. Adyen discloses no minimum-invoice figure at all — its own pricing page says only to speak with sales — and Checkout.com's enterprise-only posture means neither offers a self-serve number an SMB-volume operator can plan against. The published indicative rate card is a starting point for a sales conversation, not a contract an operator can assume they qualify for.

4. Ignoring auth-rate as a cost. A 2-percentage-point authorization-rate gap is usually larger than the entire processing-fee delta between providers. Uplift, Authorization Boost, and Intelligent Acceptance all claim ~+3.8–6%. The comparison that actually matters is landed effective cost per successful dollar, not headline rate. On a $100M annual run-rate, +2pp of auth is $2M of revenue that 30bps of fee savings will never recover.

5. Dispute economics asymmetry. Stripe's dispute pricing is two fees, not one — a received fee on every dispute, plus a countered fee refunded only on a win — cheap and predictable, but easy to under-model as a single flat $15/£20 line. Adyen and Checkout.com bury dispute pricing in the contract entirely — operators discover the real numbers post-signature, after switching costs are sunk.

6. Confusing "transparent" with "published." Both Adyen and Checkout.com market transparency. Only Adyen actually publishes indicative IC++ structure. Checkout.com publishes literally nothing on its pricing page. Marketing language and disclosed numbers are different things.

7. Comparing US-only strength to global capability. Stripe's developer-first US/UK momentum, Adyen's global enterprise omnichannel reach, and Checkout.com's EMEA/APAC digital-native focus are different kinds of strength, not points on a single ranking. Choosing a PSP for a multi-region rollout off one region's performance is the most common failure mode in international procurement — and the hardest to reverse, because integration depth varies dramatically across all three.

A Decision Framework

The question isn't which PSP is cheapest. It's which model fits the merchant profile.

Pick Stripe when: the team needs to ship in days not quarters, the product is SaaS or marketplace-shaped, embedded finance is on the roadmap, US/UK card-not-present is the primary card mix, and a published self-serve rate card is worth more than negotiating a custom contract. Stripe's blended rates plus developer experience win for merchants who haven't yet reached the scale where an IC++ negotiation pays off.

Pick Adyen when: the merchant is omnichannel (online + in-store), card mix skews enterprise (high regulated debit share, premium credit), reconciliation matters more than launch speed, and global single-acquirer coverage is operationally valuable. Adyen publishes an indicative IC++ rate card rather than a blended headline, so merchants who want interchange and scheme-fee pass-through transparency — and are prepared to negotiate volume tier discounts on top of it — tend to do better here than on a flat blended rate. European enterprise operators evaluating Adyen against a legacy-acquirer alternative should also review Adyen vs Worldpay for European Enterprise for that specific comparison.

Pick Checkout.com when: the business is digital-native and card-not-present heavy, geography centers on EMEA/APAC where Stripe and Adyen acquiring is thinner, and Intelligent Acceptance's auth-rate gains compound across high transaction volume. Checkout.com is enterprise-only by design and publishes no rate card at all — its own disclosures show 63 merchants each processing $1B+/year on its platform — so it isn't a self-serve option; it's a sales conversation from day one.

For most operators, the realistic decision is Stripe vs. Adyen — not because Checkout.com is worse, but because most operators aren't yet at the scale where an enterprise-only, quote-only PSP is worth the sales cycle. The decision between Stripe and Adyen comes down to two questions: does the merchant have negotiation leverage at IC++ scale, and does omnichannel matter more than developer velocity?

What Operators Should Actually Do

The pricing comparison that matters cannot be done from rate cards alone. It requires a card-mix model, an FX exposure model, and an auth-rate baseline measured on the merchant's own traffic.

The minimum viable comparison is: project annual volume by region; pull the BIN distribution from existing acquirer data to estimate interchange exposure; layer in dispute rate and average dispute economics; estimate FX exposure on cross-border traffic with realistic stacking; then quote each PSP with that profile in hand. The published rate card is a starting position for negotiation, not a final number.

Once volume and leverage are enough to make a custom deal worth pursuing, the highest-leverage move is requesting IC++ from Stripe and IC++ pricing breakpoints from Adyen for direct comparison. At the scale where all three vendors will negotiate, the decision is rarely about price — it's about which auth-rate optimization platform compounds best on the merchant's specific card mix.

The PSPs that win in 2026 are the ones that move auth-rate up by more than competitors move processing fees down. Pick on landed effective cost per successful dollar. The headline rate is theatre.