Embedded Finance Companies: Who Does What

TL;DR

• Embedded finance is financial services delivered inside a non-financial company's product. The provider brings the licence, rails and compliance. The platform brings the customer relationship.
• Nine companies below, grouped by what they actually embed: payments, banking, cards, connectivity, credit, buy-now-pay-later, cross-border payouts, and payroll.
• Most rosters on this topic stop at payments and cards. They leave out money going out, which is where payroll and contractor payouts sit.
• Every capability claim here links to the provider's own page. Nothing is scored, ranked or rated, because we have not run these platforms in production.
• One disclosure up front, because it shapes what you should trust: Toku is a LoudFace client. It appears in the payroll section on capability grounds, and we say so rather than hiding it.

Embedded finance is financial services delivered inside a non-financial company's product: payments, accounts, cards, credit or payouts, offered by software the customer already uses. The provider supplies the licence, the rails and the compliance work. The platform supplies the customer relationship and earns a share of the economics.

That is the whole idea. A restaurant booking tool that also advances a restaurant working capital is doing embedded finance. So is a marketplace that pays sellers into an account it issued, and a payroll system that settles a contractor in minutes rather than days.

What embedded finance actually covers

The term gets used loosely, so it helps to split it by the job being done. Each of these is a different licence, a different risk profile and usually a different vendor.

The last row is the one most lists on this topic skip, and it is not a small category. Paying people is the largest recurring outflow most companies have.

The nine companies, by what they embed

Payments and processing

Stripe

Best for: platforms that want payments, accounts, cards and lending from a single vendor rather than four.

Stripe groups four products under embedded finance: Connect for payments inside a platform, Capital for financing programmes it says can launch "in as little as five minutes", Treasury for customer accounts eligible for FDIC pass-through with ACH and wire transfers, and Issuing for virtual and physical cards that give "business owners, contractors and freelancers fast access to their earnings". Stripe cites access to 125+ payment methods. (Stripe)

The reason Stripe appears first in almost every list of this kind is breadth. One integration reaches four categories, which matters more to a small platform team than the deepest possible build of any single one.

Adyen

Best for: large platforms that want the credit risk sitting with the provider instead of on their own balance sheet.

Adyen names four: Embedded Payments, Embedded Issuing ("issue branded cards instantly" while the platform earns interchange), Embedded Accounts giving users access to settled funds, and Embedded Capital offering "instant loans up to $100k" repaid as a portion of the user's daily sales. On that last product Adyen states plainly that it "absorbs all the credit risk". It positions the stack as "purpose-built for the world's largest enterprises". (Adyen)

Who carries the credit risk is the question worth asking any embedded lending provider. Adyen answers it on the page, which is rarer than it should be.

Banking as a service

Unit

Best for: US platforms that want to launch accounts and cards in weeks rather than quarters.

Unit offers Accounts and Wallets, Money Movement across ACH, wire, cheque and real-time payments, Card Issuing for branded debit and credit, and Capital for advances and lines of credit. It reports 2 million or more users, over $100 billion in annual transaction volume, and more than 11 million daily API calls. It offers two implementation paths: a no-code "Ready-to-Launch" route it puts at three weeks, and a custom API build at six weeks. (Unit)

Read the FDIC references as a scope signal. Unit's programme is built around US bank partners, so this is a US-first choice rather than a global one.

Card issuing

Marqeta

Best for: programmes where the card itself is the product and spend controls decide the economics.

Marqeta describes itself as a card issuing platform for "debit, credit, flexible credentials, prepaid". Named capabilities include JIT Funding, dynamic spend controls, PCI-compliant widgets, virtual cards, RiskControl, and digital wallet tokenisation. It states certification to operate in 40+ countries, $400 billion in volume processed in 2025, and 99.99% platform uptime, and says customers "go live in days not months". (Marqeta)

Just-in-time funding is the detail that matters. Funds move at authorisation rather than sitting pre-loaded on a card, which changes both fraud exposure and working capital.

Financial data and connectivity

Plaid

Best for: any product that needs to read a user's bank details before it can underwrite, verify or move money.

Plaid connects more than 12,000 institutions across 20 countries. (Plaid) Its products span payments (Auth, Identity, Balance, Signal, Transfer), fraud and risk (Identity Verification, Beacon, Monitor), credit underwriting (Income & Underwriting, LendScore) and onboarding (Link, Layer). (Plaid)

Plaid is infrastructure underneath other embedded finance rather than a product a consumer meets. Most lending and account-opening flows on this page depend on something like it.

Embedded lending

Parafin

Best for: platforms that already see a merchant's sales activity and want to lend against it without becoming a lender.

Parafin offers Capital, Spend and Pay Over Time as white-labelled products, with underwriting it says is trained on over a billion cross-industry data points. It reports more than $35 billion in offers extended, over 50,000 businesses funded, and an NPS of 84. Named platform partners include DoorDash, Amazon, Walmart, TikTok, Gusto and Worldpay. (Parafin)

The partner list is the credential here. Those platforms have the sales data that makes this model work, which is also why embedded lending is hard to do without a platform relationship.

Buy now, pay later

Klarna

Best for: consumer commerce, where the payment option is also a demand channel.

Klarna offers Pay in full, Pay in 30 days, Pay in 3 or 4, and financing, alongside Sign in with Klarna, on-site messaging and express checkout. It reports 119 million shoppers, more than a million retail partners and 3.4 million daily transactions. (Klarna)

Klarna is the reminder that embedded finance is not only a B2B infrastructure story. In consumer retail it is a distribution channel that happens to be a payment method.

Cross-border accounts and payouts

Airwallex

Best for: platforms moving money out to many countries in local currency.

Airwallex exposes Connected Accounts, Accounts, Payments, Transactional FX, Payouts and Issuing through APIs. It states that platforms can accept payments in 180+ countries, transfer funds to 200+ countries, and issue local cards in 60+ countries. (Airwallex)

Note the asymmetry in those three numbers. Accepting, sending and issuing have different footprints at every provider in this category, and the smallest of the three is usually the one that constrains a launch.

Payroll and contractor payouts

This is the segment most curated comparisons on this topic leave out. ConnectPay's own guide, one of the more detailed rosters on this term, carries no payroll category at all. Broader landscape trackers file payroll separately, but the shorter, curated lists rarely build the category. That gap is also the opportunity: getting named in a vertical fintech listicle is one of the higher-ceiling levers for AI-search citation, and payroll is the category most of these lists never get to.

Toku

Best for: employers paying international teams or contractors who want settlement in minutes without replacing their payroll system.

Toku offers compliant payroll and contractor payments across "100+ countries", settled in stablecoins, with a Visa card layer so recipients can spend what they receive. Its stated integration model is additive rather than a migration: "You keep your existing system of record and add stablecoin rails underneath", connecting to ADP, Workday, Gusto, UKG and SAP through native APIs. (Toku)

Disclosure: Toku is a LoudFace client. We work on its organic and AI search visibility, and our own 30-day tracking puts Toku at 86% AI visibility, position 2.4, on its core stablecoin-payroll prompt. It is here because embedded payroll is a real category that this page would be incomplete without, and because Toku is a credible entry in it. Judge the entry on the linked source, not on our say-so. For how that visibility work shows up in practice, see the full case study.

The mechanism worth understanding, whoever you buy from: traditional cross-border payroll forces you to pre-fund and then wait through a settlement window. Faster settlement is not only an employee-experience improvement. It changes how long a finance team can keep cash deployed and how large a buffer it has to hold.

Side by side

Everything below is the provider's own published claim, taken from the pages linked above and read on 12 August 2026. We have not independently audited any of it.

Where a cell says "not stated on the page cited," that is what it means. The provider may publish the figure elsewhere. We did not go looking for a number to fill a gap, because a table that quietly mixes sourced and unsourced claims is worse than one with holes in it.

How to choose, in the order the decision actually happens

1. Name the direction the money moves. In, held, or out. Most buying mistakes start with a provider chosen for the wrong direction. Payouts and payroll providers are not interchangeable with processors.
2. Ask who holds the licence and who holds the risk. For lending, ask it explicitly. Adyen states it absorbs the credit risk. Not every provider does, and the answer changes your balance sheet.
3. Check the narrowest coverage number, not the widest. A provider that accepts in 180 countries and issues cards in 60 is a 60-country provider for anything involving a card.
4. Decide whether you are adding or replacing. Additive integrations survive procurement far more often, because nobody has to give up a system of record.
5. Then compare economics. Interchange share, FX treatment and funding mechanics differ more between these providers than headline pricing suggests.

Methodology

• What this is. A segmented reference list of embedded finance providers, grouped by the function they embed.
• How the set was chosen. We took the union of the companies named by the pages currently ranking for this topic, then added the payroll and payouts segment those pages omit. We kept the list to nine, because a tight list is more useful than an exhaustive one.
• Where every claim comes from. The provider's own live page, read on 12 August 2026, linked inline at the point of the claim.
• What we did not do. We did not run these platforms in production, so nothing here is scored, ranked or rated. There is no "number one." Claims are the vendor's, and are labelled as such.
• What would change our view. Independent implementation results. If we run a build on any of these, we will say so and update the entry.

Limitations

• Vendor-stated claims are not audited claims. Every figure above is something the provider chose to publish about itself.
• Coverage numbers move. Country counts and licence footprints change quarterly. The read date is on the table for that reason.
• The set is not exhaustive. Galileo, Railsr, Treasury Prime, Affirm, Wise Platform and others have a legitimate claim to a place here. Nine is a deliberate editorial limit. It says nothing about whether the rest matter.
• We have a client in the list. Disclosed in the summary, in the entry itself, and again here. If that context matters to how you read this, our fintech SEO and AEO work is public.