Provider directory

Everyone tells you where they operate. Almost nobody tells you which way money travels.

We reviewed 15 cross-border payment providers against their own published documentation. 3 of them state which direction value can move. For the rest, a coverage list read as though it worked both ways is wrong about half the time.

15
Providers reviewed
Against their own documentation
3
Publish direction
Collections and payouts stated separately
140
Largest claim we checked
Thunes states this many countries. Verified collection markets in this region: 2.
The directory

What they claim, and what we could verify.

Every record separates the two. Where a provider states a number and the documentation supports something narrower, both are shown — the difference is usually the useful part.

BelmoneyAggregatorChecked — not a fit

White-label remittance infrastructure sold to banks, MTOs, wallets and fintechs.

They claim

33 EEA collection countries, 130+ payout destinations.

We checked

EEA-collect to world-payout: one direction, and it collects in no SADC market. A 130-destination payout list says nothing about taking money out of Lusaka.

Payment service providerNot statedRegulator: National Bank of BelgiumChecked 2026-09-09
ConduitInfrastructureChecked — not a fit

Cross-border settlement and treasury, 15+ currencies and 100+ countries.

They claim

Used by a 40-country African network to rebalance in USDC.

We checked

Sits behind providers rather than beside them: the treasury layer a payment company uses to move its own float. Adds no leg. Evidence that the stablecoin middle leg is becoming standard infrastructure rather than a workaround.

Authorisation not establishedSame day, as documentedChecked 2026-09-09
DamisaRailTheir claim, unverified

B2B cross-border settlement, collect and pay out in fiat and stablecoins.

They claim

35+ African jurisdictions, 70+ currencies.

We checked

Names Zimbabwe as ZiG among its listed jurisdictions and publishes no direction split anywhere. Notable because every other provider that enumerates its markets EXCLUDES Zimbabwe and this one includes it. One question would either close the register's largest gap or turn a claim into a checked absence.

Authorisation not establishedSame day, as documentedChecked 2026-09-09
EversendRailTheir claim, unverified

African payment APIs: payouts, FX and a USDC off-ramp.

They claim

Ten African markets named.

We checked

Names ZMW and TZS among ten markets, with no direction split. API docs sit behind registration, so the split cannot be read without an account.

Authorisation not establishedNot statedChecked 2026-09-09
HurupayBuyer, not a supplierChecked — not a fit

Virtual USD, EUR and GBP accounts for African freelancers, backed by stablecoins.

They claim

$50m processed since January 2025.

We checked

Inbound-to-Africa only, funded by ACH and card abroad. It moves stablecoins into African payouts, which looks like a leg until you notice the direction. Adds no SADC collection. It is the buyer profile rather than a supplier.

Operates via licensed partnersNot statedChecked 2026-09-09
LocalRampRailNot established

African payouts with an institutional OTC desk above roughly $100k.

They claim

Same-day settlement from EUR and GBP into African currencies.

We checked

Covers ZMW and runs a desk above roughly $100k. Recorded as the large-ticket reference: every corridor read can price a R20,000 transfer and has nothing to say about a $150,000 one.

Authorisation not establishedSame day, as documentedChecked 2026-09-09
Mama MoneyRailNot established

Cape Town remittance operator sending from South Africa across Africa, partnered with a major SA bank.

They claim

SA-originated remittances, including to DR Congo.

We checked

Licensed and operating in South Africa, so ZAR collection is near-certain in practice. Recorded as unknown because near-certain is not checked, and this register does not seed inference.

Money transfer operatorInstant, as documentedChecked 2026-09-09
MukuruRailNot established

South African remittance operator, Zimbabwe-founded, with a large SA to Zimbabwe business.

They claim

Enterprise offering under MukuruPay.

We checked

Worth checking before anyone else on this list. They are one of the few operators with a genuine Zimbabwe presence rather than a payout relationship, so they are the most credible remaining lead on ZWG COLLECTION — the register's largest gap. No public per-direction coverage table found.

Deposit-taking institutionInstant, as documentedChecked 2026-09-09
NiumAggregatorTheir claim, unverified

Global payouts and pay-ins, with an African bank partnership reaching 35 markets.

They claim

Payouts to 190+ countries, 65 in real time. Pay-in APIs separate.

We checked

Runs separate Payout and Pay-In API families, so direction is a first-class concept for them. The per-corridor list sits behind their playbook rather than the public docs, so which African markets support which direction is not readable without access.

E-money institutionInstant, as documentedChecked 2026-09-09
OnafriqAggregatorNot established

Pan-African mobile money and bank network, collections and disbursements.

They claim

40+ countries, 1,700+ corridors claimed.

We checked

No public per-country coverage list. The developer portal renders no coverage matrix, so 40 countries and 1,700 corridors are stated and not checkable.

Payment service providerInstant, as documentedChecked 2026-09-09
pawaPayRailVerified from their documentation

Mobile money aggregator across 20 African markets, covering roughly 85% of African mobile money.

They claim

650 million active mobile wallets reached.

We checked

Publishes direction at the level of the individual mobile-money OPERATOR, which is finer than anyone else found. Mozambique proves why that matters: Vodacom takes deposits and Movitel does not, in the same country. Their table also established the first collection leg on record for Lesotho.

Collects LSL ZMW MWK TZS MZNPays out LSL ZMW MWK TZS MZN
Operates via licensed partnersInstant, as documentedChecked 2026-09-09
PretiumRailTheir claim, unverified

Stablecoin payments across African markets, settling on Celo, Base and BNB.

They claim

7+ African markets.

We checked

Names ZMW, MWK and CDF among seven markets and publishes no direction split. Public API docs exist and may enumerate it.

Authorisation not establishedInstant, as documentedChecked 2026-09-09
TerraPayAggregatorNot established

Money movement network focused on wallets and mobile money across many corridors.

They claim

Bank, wallet and mobile money interoperability.

We checked

Their network page returns 403 to a plain fetch, so nothing has been established.

Payment service providerNot statedChecked 2026-09-09
ThunesAggregatorVerified from their documentation

The largest payout network encountered: 140 countries, 90 currencies, 720 connected members.

They claim

140 countries reached, 220 payment methods.

We checked

Their site separates a Pay Network from an Accept Network. Accept is collections, and in Africa it covers exactly two countries: Tanzania and South Africa. The Pay Network reaches around thirty-eight African markets including Zimbabwe. Reading the 140-country headline as bidirectional is wrong by a factor of nineteen on this continent.

Collects ZAR TZSPays out ZAR TZS ZMW MWK MZN BWP LSL ZWG
Payment service providerInstant, as documentedChecked 2026-09-09
Yellow CardRailVerified from their documentation

Licensed across ~20 African markets, with Visa partnership for stablecoin settlement.

They claim

20 African countries with on and off ramp.

We checked

The only provider found that publishes Receives and Sends as two separate country tables. That is how Botswana's collection leg was established. Zimbabwe and Namibia appear on neither table.

Collects BWP MWK ZMW TZS ZARPays out BWP MWK ZMW TZS ZAR
Crypto-asset service providerInstant, as documentedChecked 2026-09-09
How the market divides

Counts, not percentages.

Fifteen providers is not a sample worth taking a percentage of. Unknown is shown rather than omitted, because on most of these axes it is the largest group and hiding it would overstate how much is actually established.

By what they are

Rail
8
Aggregator
5
Infrastructure
1
Buyer, not a supplier
1

By authorisation

Authorisation not established
5
Payment service provider
4
Operates via licensed partners
2
Money transfer operator
1
Deposit-taking institution
1
E-money institution
1
Crypto-asset service provider
1

By documented speed

Instant, as documented
8
Not stated
4
Same day, as documented
3

Regulatory class is what the provider operates under, where it is stated. Settlement speed is what they document, not what we measured — nothing here has been timed end to end. `unknown` counts are shown rather than omitted, because on most axes they are the largest group.

For developers and agents

The same directory, as data.

No key required. Filterable by what a provider is, what it is authorised as, and how fast it documents settlement.

GET /api/v1/providers
GET /api/v1/providers?regulatoryClass=psp
GET /api/v1/providers?currency=ZMW
GET /api/v1/providers/{id}

See which routes actually work →

Know a provider we have not checked?

The directory grows by reading documentation, not by taking claims on trust. If you work with someone covering a Southern African corridor, tell us and we will check them and publish what we find.