"Stripe charges 2.9% + 30¢" is the most-repeated and least-useful sentence ever written about the company. It's true — and it explains almost nothing about how Stripe actually makes money in 2026 or why they're worth ~$70 billion. The interesting business is happening one layer deeper.
This piece is the developer-focused breakdown of Stripe's real economics: what they actually keep from that 2.9%, how the layered products (Tax, Billing, Connect, Issuing, Atlas, Capital, Treasury) became the actual margin business, and what that means strategically for any company building on top of them.
The 2.9% + 30¢ — where it actually goes
When a US customer pays you $100 with a credit card, you see Stripe charge $2.90 + $0.30 = $3.20. Stripe does not keep $3.20. The breakdown looks roughly like:
| Bucket |
Goes to |
Approximate share of $3.20 |
| Interchange |
The customer's card-issuing bank (Chase, Citi, etc.) |
~$1.80 |
| Network fees |
Visa / Mastercard / Amex |
~$0.20 |
| Acquirer / processing partner fees |
Bank/network plumbing |
~$0.30 |
| Stripe gross take |
Stripe |
~$0.90 |
Numbers vary by card type (rewards cards have higher interchange), region, and merchant category. But the rough split: Stripe keeps somewhere between 25%–35% of the headline 2.9% — call it 0.3% of payment volume, in the ballpark.
That gross take then has to cover their own costs (engineering, fraud losses, support, compliance, marketing). So the actual net margin on raw card processing is, structurally, modest. This is why Stripe's competitors who try to undercut on raw rates (e.g. some "1% flat" pitches) tend to either go bankrupt, switch business models, or specialise in markets where interchange is cheaper.
If Stripe were only a payment processor in 2026, they'd be a perfectly good ~$10–15B company. They're worth ~$70B because they're not just a payment processor anymore.
The layered products — where the actual margin lives
Starting around 2018 and accelerating dramatically by 2026, Stripe has built a stack of higher-margin products that bolt onto the core processing rail. Each one solves a real customer problem — and each one collects an additional fee on the same payment volume.
Stripe Billing (~0.5% of subscription revenue, free up to $1M/yr)
The subscription management layer. Handles recurring charges, proration, dunning, tax-aware invoicing, customer portals. Replaces what used to be 6 weeks of internal engineering. Pure software margin (~80%+) on top of payment processing.
Stripe Tax (0.5% of transactions)
Calculates and collects sales tax / VAT / GST in 50+ jurisdictions. Doesn't replace tax filing — but automates the calculation. Half a percent on a $1B-volume merchant is $5M/year. Pure software margin again.
Stripe Connect (varies, often 0.25%–1% on volume)
The marketplace-and-platform layer. Powers Shopify, Lyft, DoorDash, Substack, Amazon-like multi-seller flows. KYC, payouts to sub-merchants, split payments — all of it. This is the moat product. Once a marketplace is on Connect, switching is a multi-quarter re-platform.
Stripe Issuing
Lets companies issue their own physical and virtual cards. Used by Ramp, Brex, Shopify Balance, and any "card on us" workflow. Stripe earns a slice of interchange on each card swipe — and these are programmable cards (controls, vendor lock, expense rules), so the data is incredibly valuable.
Stripe Capital
Cash-advance product for Stripe merchants. Stripe sees your revenue history (because it's flowing through Stripe), pre-approves you for a cash advance, and you repay automatically as % of future sales. Higher effective rates than bank loans, but instant approval and zero paperwork. Very profitable book of business.
Stripe Atlas
Incorporates a Delaware C-Corp for $500 (one-time) plus optional $100/year. Becomes the on-ramp for international founders setting up US companies — and once you've used Atlas, you're using Stripe payments by default. Atlas isn't a profit center directly; it's a customer acquisition channel for the rest of the stack.
Stripe Treasury
Banking-as-a-service for platforms that want to offer their own users a bank account. Powers Shopify Balance, etc. Stripe earns a slice of the deposits + a usage fee.
Stripe Climate
Carbon removal as a checkout option. Less directly profitable, more brand and regulatory positioning. But it gets Stripe in front of every CFO who cares about ESG reporting.
The strategic thesis: financial infrastructure, not payments
Add it all up and Stripe in 2026 looks less like "PayPal for developers" and more like the financial infrastructure layer for the next generation of internet businesses. The 2.9% + 30¢ is the acquisition channel — it gets developers and startups in the door. The layered products are where the long-term margin lives.
This is why their valuation reflects something closer to a software platform than a payment processor. Adobe, Salesforce, ServiceNow trade at higher revenue multiples because their gross margins are higher and their lock-in is stronger. Stripe is moving toward that profile.
The 2025 numbers (publicly available context)
Stripe is private, so we don't get full GAAP statements. But what they've shared publicly:
- Total payment volume (TPV) in 2024: ~$1.4 trillion (reported by Stripe in their 2024 annual update)
- Net revenue (the part Stripe keeps): roughly proportional — likely $20B+ on that volume
- Profitability: they hit cash-flow positive in 2024 and have stayed there
- Last private valuation: ~$70B in 2024 secondary market, with chatter about an IPO in 2026 or 2027
- Customer base: millions of businesses, including ~78% of the Forbes Cloud 100
For comparison: Adyen (Stripe's main public-market peer) processes about 60% of Stripe's volume and trades at ~$45B. Stripe's premium reflects both growth rate and the wider product stack.
What this means for developers building on Stripe
Three concrete implications:
1. Stripe is unusually motivated to keep developers happy
The platform-level products (Connect, Issuing, Treasury) only generate revenue if developers build on them. So the documentation, the SDKs, the testing tools, the dashboard, and the support are all unusually good — because that's the moat. This is rare in fintech and worth appreciating.
2. Lock-in compounds across products
Switching off Stripe Payments alone is annoying but doable — your subscription database, your customer payment methods, and your webhook history all migrate with effort. Switching off Stripe Payments + Tax + Connect + Atlas is a six-month engineering project that almost no company actually does.
This isn't sinister — it's just how platform pricing works. But it means: be thoughtful about which layered products you adopt. Each one is a one-way door. If you only need payments today, only use payments. Add Tax when you actually have a tax problem. Add Connect when you're actually a marketplace.
3. The $0.30 fixed fee is the under-appreciated tax on small transactions
For a $5 SaaS subscription, the $0.30 fixed fee is 6% — bringing your effective rate to ~9%. For a $50 transaction, it's negligible. This is why "$1/month" pricing strategies are often economically unworkable on Stripe — the fixed fee eats your unit economics. Either price higher, or pick a processor with a smaller fixed fee for micro-transactions (e.g. some specialised micropayment platforms, though they have other tradeoffs).
What Stripe doesn't make money on (and why it matters)
A few things Stripe gives away or near-gives-away — and the strategic reason why:
- The Customer Portal. Free with any Stripe account. Saves you from building cancel/upgrade/invoice flows yourself. Strategic reason: makes Stripe sticky vs competitors who charge for portal.
- Stripe Docs. The best technical documentation in fintech, free. Strategic reason: developer love drives bottom-up adoption.
- Stripe Sigma & Data Pipeline (paid, but cheap). Lets you query your own Stripe data in your own warehouse. Strategic reason: deepens platform lock-in.
- Test mode and the Stripe CLI. Genuinely free, genuinely powerful. Strategic reason: lowers the activation friction to near-zero.
The pattern: anything that makes you successful on Stripe is free or cheap. Anything that makes Stripe successful through you (Tax, Connect, Issuing) is the revenue.
Compared to competitors
Briefly, where Stripe sits vs the alternatives:
- Adyen — direct competitor, similar layered model, more enterprise-flavoured. Lower fees at scale, less indie-friendly.
- Lemon Squeezy / Paddle — merchant of record (we covered this in Best payment gateways for SaaS 2026). Charge ~5% but handle global tax. Different business model — sell-through, not pass-through.
- Block (Square + Cash App) — strongest in physical retail and consumer; not in the same lane for SaaS.
- PayPal / Braintree — same headline rates as Stripe, much weaker developer experience and product stack. Used by businesses with consumer-trust requirements.
Stripe's position isn't that they're the cheapest — they're not. It's that they're the most complete financial platform for businesses that ship code.
The IPO question (a brief aside)
Stripe has been "about to IPO" for ~5 years. As of April 2026 they remain private, employees have done multiple secondary share sales, and the public market would clearly love them. If they do go public, two things to watch:
- Take rate trends — are they keeping more or less of payment volume over time? Margins are the story.
- Layered-product revenue mix — what % of revenue comes from Tax, Billing, Connect, Issuing, etc.? The higher this number, the more they look like a software company instead of a payment processor — and the higher their public multiple should be.
What devs can take from this
A few practical takeaways:
- Build on Stripe with eyes open about what you're adopting. Payments-only is portable. Payments + Tax + Connect + Atlas is a relationship.
- Use the free tools aggressively. The Customer Portal alone saves weeks of engineering. The Sigma queries are cheaper than building the same in-house.
- Watch the fixed-fee math on small transactions. Adjust pricing accordingly or pick a different rail for micropayments.
- Treat Stripe Capital as a credit product, not a free-money fountain. The effective rates are higher than they look — fine for working capital, painful for long-term debt.
- If you're shipping a marketplace, Connect's lock-in is real. Pick deliberately.
FAQ
Is Stripe actually profitable in 2026?
Yes — they were cash-flow positive starting in 2024 and have stayed there. Their total profit margins are still modest because they reinvest aggressively in product, but the unit economics work.
Why is Stripe more expensive than my old merchant account?
Because that old merchant account had hidden fees (PCI compliance fees, statement fees, monthly minimums, gateway fees, monthly subscription) that added up to more than 2.9% + 30¢ in practice. Stripe consolidated all of that into one transparent rate.
How does Stripe make money on free Customer Portal usage?
They don't, directly. The Portal is loss-leader infrastructure to keep you on Stripe. The math works because once you're on Stripe Billing, you're paying the 0.5% on subscription revenue.
Is Stripe Tax actually worth 0.5%?
For most US-only businesses with mostly straightforward sales tax: probably not — a CPA + their existing dashboard handles it. For businesses selling globally to consumers (where you'd otherwise need to register for VAT/GST in dozens of countries): yes, easily.
Will Stripe lose to a cheaper challenger?
Unlikely on price alone. Stripe's moat is the integrated product stack + developer experience, not the headline rate. Pure-price challengers tend to either go bankrupt or get acquired.
How much do Visa and Mastercard make on every Stripe transaction?
Their combined network fees on a typical US transaction are ~$0.20 per $100, or ~0.2% of volume. They're a duopoly with extraordinary margins — payment networks are arguably better businesses than Stripe.
Could a regulator break Stripe's lock-in?
Possible. The EU and UK have already taken interest in payment-fee structures. Most likely outcome over the next few years is more transparency requirements (showing customers exactly what fees go where), not forced break-up.
The bottom line
Stripe makes most of its money the way most modern platforms do: by charging a small toll on a giant base of transactions, and selling higher-margin software products to the businesses already on the toll road. The 2.9% + 30¢ is the toll. Tax, Billing, Connect, Issuing, Atlas, Treasury are the actual business. Understanding that distinction makes you a better operator on top of Stripe — and a better consumer of fintech analyst takes that confuse "payment volume" with "revenue".
Related reading