The press releases landed within hours of each other on a late February morning in 2026. Record revenues. Record profits. Record customer growth. At first glance, Chime and Nubank appeared to be reading from the same script—challenger banks that had finally cracked the code on profitability.
Look closer, though, and the similarities evaporate.
Chime reported $2.2 billion in revenue for 2025, a 31% year-over-year jump, with roughly 9.5 million active members each generating about $257 annually. The San Francisco-based company said it expects to post its first full-year GAAP profit in 2026, eight months after a June IPO that valued it around $11 billion. Nubank's numbers operated in a different orbit entirely: $16.3 billion in revenue for the same period—up 45%—and $2.9 billion in net income from 131 million customers scattered across Brazil, Mexico, and Colombia. Return on equity: 33%.
The gap isn't merely one of scale. It's philosophical. Chime is building a high-margin liquidity engine atop proprietary infrastructure while maintaining a partner-bank model that keeps regulatory burdens at arm's length. Nubank is doing the opposite—aggressively securing full banking licenses, owning the balance sheet outright, and assembling a sprawling ecosystem that stretches from crypto to insurance.
Both believe they've solved the existential riddle that has plagued digital banking for a decade: how to make money without relying on interchange fees alone. Whether either strategy proves durable is another question.
When Free Checking Stops Being Enough
The early neobank pitch was seductive in its simplicity. Free checking accounts. High-yield savings. Direct deposits that hit two days early. Mobile interfaces that made Bank of America's app look like it was designed on a Motorola Razr. Revenue? Interchange fees—those small cuts processors collect every time someone swipes a debit card.
It worked brilliantly for user acquisition. It worked terribly for margins.
By 2024, the pivot was underway. Chime launched Chime+ in March 2025, a premium membership tier that bundled extra rewards, higher APY, and exclusive perks—unlocked, notably, by qualifying direct deposits. A redesigned secured credit card followed in September, its cash-back rewards woven into the Chime+ ecosystem.
The more significant move, though, was invisible to most customers. Buried in Chime's February 2026 earnings release was a single line: the company had completed its migration to ChimeCore, its proprietary banking core, by the fourth quarter of 2025. Processing costs dropped roughly 60%. Product development cycles shortened. It's the kind of multi-year infrastructure bet that sounds tedious in investor presentations but fundamentally rewires unit economics.
Platform-related revenue—the company's term for anything that isn't interchange—tells the rest of the story. Chime generated $200 million from these sources in Q4 2025 alone, up 47% year-over-year. For the full year, platform revenue hit $686 million, a 73% increase driven largely by MyPay, the earned wage access product launched back in May 2024. By late 2025, MyPay was running at an annualized revenue pace exceeding $400 million, with transaction margins around 60% and loss rates near 1%.
Those are numbers that make venture capitalists lean forward in their chairs.
Liquidity Products, or How to Escape the Interchange Trap
If interchange was the neobank playbook from 2015 through 2020, liquidity products have become the growth lever of choice in 2025 and beyond. Chime's suite—SpotMe (fee-free overdraft), MyPay (earned wage access), and Instant Loans (three-month installment loans for pre-approved members)—collectively originated more than $40 billion on an annualized basis as 2025 ended. By Q4, one in ten active Chime members carried an open Instant Loan.
This isn't payday lending with a millennial-friendly UI. Chime's loss rates remain low, in part because the company underwrites using proprietary data: direct deposit relationships, spending patterns, transaction velocity. The Credit Builder secured card, introduced in 2020, helped establish credit files for customers who might otherwise remain invisible to traditional lenders. MyPay, initially marketed as a safety net for workers caught between pay periods, evolved into something more lucrative—a high-margin recurring revenue stream with better unit economics than interchange ever delivered.
Then there's the enterprise angle. In March 2025, Chime rolled out Chime Enterprise and Chime Workplace, a financial wellness suite distributed through employers. Partnerships with Workday (announced in August) and UKG (November) turned payroll platforms into customer acquisition engines. Employer-sponsored financial wellness isn't exactly novel. But Chime's version integrates liquidity products directly into the flow of work—advances before payday, automated savings from direct deposits, credit-building tools accessible without leaving an HR portal.
It's banking as an employee benefit. Whether that scales beyond early adopters remains to be seen.
Nubank's License Obsession
Nubank placed a different wager: if you want to be a bank, become one.
In April 2025, Mexico's National Banking and Securities Commission approved Nu Mexico's banking license, allowing the company to graduate from a limited-license Sofipo to full banking operations. By February 2026, Nu Mexico had roughly 14 million customers—about 15% of Mexico's adult population—and the company committed to investing $4.2 billion in the market through 2030.
The license matters because it eliminates the margin drag inherent in partner arrangements. Nubank had already accumulated $41.9 billion in deposits globally by year-end 2025, up 29%, alongside a credit portfolio of $32.7 billion, up 40%. Owning the balance sheet outright means capturing both sides of the spread: paying competitive rates on "Cajitas" (goal-based savings accounts) while deploying capital into credit cards, personal loans, and payroll-linked lending.
Then came the United States.
On January 29, 2026, the Office of the Comptroller of the Currency granted conditional approval for Nubank to establish a de novo national bank—Nubank, N.A. The application, first announced in September 2025, positions the company to operate in the world's largest financial market with full regulatory parity. It's a long-term play, and not without risk. De novo charters bring capital requirements, compliance burdens, and multi-year buildout timelines. But it signals something about Nubank's strategic patience: the company is willing to absorb complexity in exchange for structural advantage.
Not every fintech has that appetite.
The Everything Strategy
While Chime narrows its focus on liquidity and infrastructure efficiency, Nubank is assembling what can only be described as a financial superapp. The product catalog is sprawling: credit cards, high-yield savings, investment accounts (NuInvest), insurance (the company had roughly 2 million active policies in Brazil as of mid-2024 through a partnership with Chubb), personal loans, SME accounts, crypto trading (NuCripto), and even a proprietary loyalty token (Nucoin, launched in March 2023).
In Colombia, Nubank introduced CDT Nu—term deposits starting at 50,000 pesos with 90- to 120-day terms offering rates as high as 11.1%—in February 2025. In Brazil, the company partners with Circle to offer USDC and with Lightspark to integrate Lightning Network payments. These aren't vanity projects. They're infrastructure bets on programmable money rails that could someday power cross-border remittances at a fraction of traditional FX costs.
Maybe.
Revenue per customer reflects this expanding product surface. Nubank's average revenue per active customer (ARPAC) hit $15 in Q4 2025, roughly double the $7 to $8 range reported in 2021. The efficiency ratio—operating expenses as a percentage of revenue—stood at 19.9%. Cost-to-serve dropped to $0.80 per customer. That combination of rising monetization and falling unit costs is what sustainable profitability looks like when it actually works.
Brazil remains the anchor: 113 million customers by late February 2026, representing 62% of the adult population. Mexico is the growth engine. Colombia, with 4 million customers (roughly 11% of adults), serves as the testing ground for replicating the playbook in smaller, more fragmented markets.
AI as Infrastructure, Not Marketing
Both companies emphasize artificial intelligence, though to different ends. Chime deployed generative AI voice and chat bots throughout 2025, claiming improved customer satisfaction and lower service costs. The company's cost-to-serve has fallen roughly 30% over three years. AI isn't the only driver, obviously—process optimization and automation play roles—but handling customer service at scale without human escalation requires machine learning models capable of navigating nuance.
Nubank's AI work cuts deeper into the product layer. The company developed nuFormer, a proprietary underwriting model that ingests transaction data, social signals, and behavioral patterns to extend credit to customers traditional banks would summarily reject. Nubank's PIX integration—Brazil's instant payment system—uses AI to detect fraud in real time, with more than 10 million monthly active users interacting with AI-driven features.
This isn't about efficiency alone. It's about unlocking markets. When you can profitably underwrite a $500 loan, you can serve populations legacy institutions ignore. When fraud models adapt faster than bad actors, you can move money instantly without bleeding losses. AI becomes the difference between a neobank that resembles a cost-optimized version of Wells Fargo and one operating in an entirely different risk-reward envelope.
What the Numbers Reveal
The 2025 results weren't flukes—or at least, they don't appear to be. Chime's path to GAAP profitability in 2026 rests on platform-related revenue outpacing payments revenue, ChimeCore driving down variable costs, and enterprise distribution scaling without proportional customer acquisition cost increases. The June 2025 IPO at $27 per share raised roughly $800 million and imposed public-market discipline. Investors will be watching whether ARPAM can climb beyond $257 as Instant Loans and premium tiers mature.
Nubank's $2.9 billion in net income wasn't a one-time margin expansion trick. The company has been profitable since 2021, and its 33% ROE suggests it's not sacrificing returns for growth. The Mexico and U.S. licenses are multi-year bets, granted, but they position Nubank to operate in markets where regulatory arbitrage is narrowing. The $4.2 billion Mexico investment signals confidence that the Brazilian playbook—interchange as customer acquisition, then layer in deposits, credit, and adjacent services—translates across borders.
Whether that confidence is warranted is another matter.
Risks Neither Company Likes to Discuss
Both strategies carry embedded vulnerabilities. Chime's partner-bank model insulates it from certain regulatory burdens but limits balance-sheet flexibility. If deposit rates remain elevated or credit losses spike in a downturn, partner banks may reprice relationships—or exit them entirely. Nubank's license strategy brings capital requirements and compliance overhead that could slow product iteration. The U.S. charter, in particular, subjects the company to OCC supervision at a moment when regulators are scrutinizing fintechs with renewed intensity.
There's also the question of replicability. Chime's proprietary core and liquidity products provide structural cost advantages, but those products aren't defensible indefinitely. Traditional banks can—and some already do—offer earned wage access. Fintech competitors can build installment lending engines. The real moat may be the direct deposit relationship itself, that moment when a customer makes Chime their primary account. But that's a behavioral moat, not a technical one, and behavioral moats can erode.
Nubank's licenses and scale in Latin America create regulatory and distribution advantages that are genuinely difficult to replicate. A startup can't wake up tomorrow with 131 million customers and a Brazilian banking license. But the ecosystem strategy requires constant reinvestment in new products, and there's no guarantee insurance or crypto or wealth management will drive meaningful ARPAC gains. The risk is becoming a conglomerate without focus, a financial services Frankenstein's monster stitched together from disparate parts.
Two Answers to the Same Question
The narrative that neobanks can't make money aged poorly. Chime and Nubank proved the model works—just not the same model. One built a high-margin liquidity engine on someone else's balance sheet. The other became the balance sheet and wrapped it in an ecosystem of ancillary products designed to turn banking into a platform.
Perhaps the more revealing question is whether either approach creates a durable competitive advantage. Markets have a tendency to arbitrage away first-mover benefits. What looks like a structural edge in 2026 can become table stakes by 2028.
What both companies undeniably share is a refusal to remain trapped by interchange economics. They saw the margin compression coming years before most competitors and made different billion-dollar bets on what comes next. By February 2026, both bets were paying off, at least in headline numbers.
The harder test arrives later—in 2030, when the next wave of challengers emerges with business models Chime and Nubank didn't anticipate, and when the question shifts from "Can neobanks be profitable?" to "Can today's profitable neobanks stay that way?"
That's a story still being written.
