Press send on a $200 transfer to sub-Saharan Africa and, on the World Bank’s latest count, $16.92 of it never arrives. That is 8.46 percent, the Q3 2025 average, the highest of any receiving region and well above the global 6.36 percent. Crypto remittances to Africa are sold as the way around that bill: turn dollars into a stablecoin, push them across a public chain for a few cents, and let someone at the far end pay out in naira or shillings.
The chain part is true. It is also the smallest part. A stablecoin transfer is cheap in the middle and expensive at its two doors, the place where dollars become tokens and the place where tokens become money a mother in Nairobi can spend. This lesson prices every leg of that $200, names the companies that run the rails, and shows when the route wins and when it loses.
It follows our lesson on stablecoins as dollar access in emerging markets, which covers depegs, issuer freezes and savings, so none of that is repeated here. If tokens themselves are new to you, start with what stablecoins are and how they hold their peg. For where this fits among other uses of crypto on the continent, see our overview of DeFi in Africa: a remittance rarely touches DeFi at all, and the honest label for most of what follows is payments on a public chain.
What sending $200 to Africa costs without crypto
The World Bank’s Remittance Prices Worldwide is the reference. Its Issue 54, dated September 2025 and covering Q3 2025, was still the latest listed on the project’s homepage when we checked on 2 November 2026. It measures the total cost of sending $200, meaning the transfer fee plus the margin hidden in the exchange rate, across 358 country corridors. The averages below are that total, and the provider you pick matters more than the region.
| Route or region (cost of sending $200) | Q3 2025 | Note |
|---|---|---|
| Global average | 6.36% | 6.49% in Q1 2025 |
| Sub-Saharan Africa, receiving | 8.46% | 8.78% in Q1 2025; most expensive region |
| Middle East, North Africa, Afghanistan, Pakistan | 5.11% | Cheapest receiving region |
| Banks | 14.99% | Most expensive provider type |
| International MTO Index (Western Union, MoneyGram) | 5.52% | 5.91% in Q1 2025 |
| Digital remittances index | 4.59% | Sent online, received in an account or wallet |
| Digital-only MTO index (Wise, Remitly and others) | 3.54% | Best benchmark for a crypto route |
| Paid out to a mobile wallet | 3.92% | 481 services |
| From the US, all destinations | 5.04% | From the UK: 4.61% |
| Target for 2030 (UN goal) | 3% | Global average; no corridor above 5% |
Two things in that table change how to read any crypto pitch. First, 8.46 percent is an average across corridors, not a quote, and it includes services nobody cost-conscious would choose. Comparing a stablecoin route with it flatters the stablecoin. The fairer rival is the 3.54 percent charged by digital-only firms. Second, banks at 14.99 percent drag up every average they sit in.
The site lets you select a sending and receiving country and see prices by provider, which beats any regional number. South Africa shows why that matters: as a sending country it costs 15.65 percent on average, the dearest of the G20, so moving money within Africa is often worse than moving it from Houston or Manchester.
Mobile money has already cut the bill
Crypto is not the only challenger. The GSMA reports that $45 billion of international remittances ran through mobile money in 2025, up 23 percent, across 381 million transfers, and that sub-Saharan Africa accounts for three quarters of that value. It puts the cost of sending $200 by mobile money at 3.63 percent in Q1 2025, using World Bank data.
The World Bank’s own number for mobile operators is less tidy. Its Q3 2025 report records 5.58 percent, on less than 1 percent of its sample, and explains that the operators’ cost hovered around 3 percent for years with fees of 1.5 to 2 percent, the swings coming largely from the exchange margin. Treat it as a thin reading. The lesson in that detail applies to crypto too: in unstable currencies, the rate moves the total more than the fee does.
How big crypto remittances to Africa are today
Chainalysis published the most recent figures on 29 October 2026, for the year to June 2026. Western countries sent $888 million of crypto remittances to the region, up 36 percent, in more than one million transfers. North America accounted for $645 million and Western Europe for $100 million. The United States sent $223 million, France only $5.4 million, though France hosts nearly as many people born in the region. Stablecoin remittances to the region, a separate count, more than doubled from $235 million to $492 million while bitcoin’s stayed largely flat.
Set that against the total. Chainalysis itself puts remittance inflows to the region at $41 billion in 2025, while a World Bank brief of June 2024 counted $54 billion in 2023. On either base, crypto comes to about 2 percent of the flow. Chainalysis does not explain how it defines a remittance, and informal transfers escape both counts, so take the proportion as rough. An implied average of under $900 per transfer also suggests these are not the typical $200 sends the World Bank prices.
Four legs, not one: how a stablecoin transfer is built
A bank wire looks like a single price. A stablecoin transfer is four separate deals, often with four different firms. Our guide to how to buy crypto in Africa covers ramps from the buyer’s side; here is how they chain together for someone sending money home.
Leg 1: the on-ramp in the sender’s country
You buy USDC or USDT with dollars on an exchange or app, after identity checks. Do not assume which payment method is cheapest: among ordinary remittance firms the World Bank found cards at 4.39 percent and bank accounts at 8.69 percent in Q3 2025, and exchanges price their own way. The cost shows up in three places: a stated fee, a spread between the buy price and a dollar, and sometimes a withdrawal fee when you move the tokens out. Add them. An exchange advertising zero fees can still charge through the spread.
Leg 2: the network fee
Marketing talks about this leg, and the numbers depend on the chain. Just over half of all USDT lives on Tron: DefiLlama showed 92.3 billion of the 179.2 billion dollars in circulation there on 2 November 2026, against 2.4 billion on Solana. USDC is at 6.8 billion on Solana and 4.4 billion on Base.
Tron is not the cheapest. Its chain parameters on 2 November 2026 priced energy at 100 sun a unit. A USDT transfer takes about 65,000 energy, or about 130,000 if the receiving address holds no USDT yet: two simulated transfers on a public Tron node that morning returned 64,285 and 130,285. That burns about 6.5 or 13 TRX for a wallet with no staked energy. With TRX near $0.321 at 10:08 UTC, the cost is about $2.09 or $4.17, an estimate. On Base, gas read at about 0.006 gwei the same morning, which puts a token transfer at a cent or less before any data fee. On Solana the base fee is a fraction of a cent.
A flat fee hurts small amounts most. $2.09 is about 1 percent of $200, about 4 percent of $50 and about 21 percent of $10. For small and frequent transfers, the choice of chain decides whether the route makes sense at all, and sender and receiver must both be on the same one: USDT sent on the wrong network can be lost.
Leg 3: the off-ramp and its spread
Someone must hand over naira, shillings or cedis in exchange for the tokens. That someone quotes a rate, and the gap between that rate and the mid-market rate is the real price of the transfer. The World Bank’s survey covers banks, money transfer firms, mobile operators and post offices, not crypto ramps, so the spreads in the next section are assumptions. What is documented is the mechanism: where official and street rates split, as the World Bank describes for Malawi, margins jump from one quarter to the next.
Watch the reference rate too. If the off-ramp pays above the official rate, the recipient may feel ahead, but the comparison that counts is with what a licensed remittance firm would deliver on the same day, in the recipient’s hand.
Leg 4: the last mile
The recipient rarely wants a wallet address. They want a credit on M-Pesa, MTN MoMo or Orange Money, or a bank balance. Getting there can add a payout fee, and then spending the money or withdrawing it from an agent has the ordinary tariff of that wallet. A traditional transfer that ends in the same wallet pays that final charge as well, so it does not count against crypto, but it does belong in the sum of what reaches the family.
Crypto remittances to Africa worked through: $200 against the averages
What follows is an example, not a quote from any provider. The World Bank figures are real. Network fees are the estimates above, dated 2 November 2026. The on-ramp costs and off-ramp spreads are our assumptions: we test a range of 0 to 3 percent going in and 2 to 8 percent coming out, to show how far the result moves. Each percentage applies to what is left at that leg, and any last-mile fee is left out on both sides.
| Example (assumed ramps) | On-ramp | Network | Off-ramp spread | Total cost | Value delivered |
|---|---|---|---|---|---|
| A: already holds stablecoins | 0% (not bought for this) = $0.00 | Base or Solana: $0.01 | 2% = $4.00 | $4.01 (2.0%) | $195.99 |
| B: cheap ramps | 1% = $2.00 | Base or Solana: $0.01 | 2% = $3.96 | $5.97 (3.0%) | $194.03 |
| C: middling ramps | 2% = $4.00 | Tron, receiver holds USDT: $2.09 | 4% = $7.76 | $13.85 (6.9%) | $186.15 |
| D: dear ramps | 3% = $6.00 | Tron, empty receiver: $4.17 | 8% = $15.19 | $25.36 (12.7%) | $174.64 |
| Real benchmark, Q3 2025 | Cost | Cost on $200 | Value delivered |
|---|---|---|---|
| Banks (World Bank) | 14.99% | $29.98 | $170.02 |
| Sub-Saharan Africa average (World Bank) | 8.46% | $16.92 | $183.08 |
| Digital-only MTO index (World Bank) | 3.54% | $7.08 | $192.92 |
The network fee hardly matters at $200. Assumed spread does almost all the work: in scenario D the off-ramp alone costs $15.19, about 60 percent of the whole bill.
Scenario B beats the regional average by about $11, yet beats a digital-only firm by only $1.11, before you count the time spent, the risk of a mistake and the record-keeping. Scenario C loses to the digital firm by more than $6. Scenario D loses to the regional average as well. Only the person who already holds stablecoins, in scenario A, gets a clear saving, and that person is a freelancer paid in USDC more often than a migrant paid in dollars at the end of the month.
To run this for your own corridor, take a live quote from a licensed transfer service and note how much local currency arrives. Then get the off-ramp’s quote for the same moment and work backwards through the fees. Local currency delivered is the only figure that compares like with like.
The companies behind the rails, and who they serve now
The names that come up in every article on crypto remittances to Africa are not the consumer apps they once were, or were assumed to be. The table describes each one as its own website did when read on 2 November 2026. These are companies, not endorsements, and prices depend on the corridor and the client.
| Company | What it says it does | Reach, per its site | Who it serves |
|---|---|---|---|
| Yellow Card | Digital asset and fiat infrastructure: API, treasury portal, global accounts | 60+ countries | Banks, financial institutions, corporates |
| Bitnob | Global payouts, multi-currency wallet (NGN, USDT, USDC, BTC), card issuing | Payouts to 12+ countries, not listed | Businesses and developers |
| Kotani Pay | Cash-in and cash-out for digital assets, APIs, an SMS-based wallet | “Across Africa”, countries not listed | Businesses and dapps; end users through the SMS wallet |
| Fonbnk | Stablecoin settlement from prepaid payments, with a sell flow to local rails | 19 markets, 15 of them with mobile money rails | Businesses, plus a consumer sell flow |
| Strike | Bitcoin buying and payments | Over 100 countries, not listed | Mainly individuals |
| Machankura | Bitcoin over USSD and SMS, tied to a phone number | Ten African countries | People on any phone |
Yellow Card leaves retail
Yellow Card spent years selling stablecoins to individuals through its app. TechCabal reported that it emailed users on 29 October 2025 telling them to withdraw before 31 December, and that from 1 January 2026 it would work only with institutions; Condia gave the same closing date for the app. At a briefing in Lagos reported on 14 August 2026, Yellow Card Nigeria head Lasbery Oludimu called it “a business and product-related decision” and denied any link to Binance’s troubles in Nigeria. Compliance costs were a key factor, he said: every customer onboarded triggers paid checks, and the firm judged its infrastructure earned more sold to companies.
On 7 May 2026 Mastercard and Yellow Card announced a partnership covering Ghana, Kenya, Nigeria, South Africa and the United Arab Emirates, with cross-border remittances among four target uses. The announcement describes working groups and pilots, not a live product. So anyone who searches for Yellow Card as a personal app today will not find one, and any site offering it should be treated with suspicion.
Why most of the market went business-to-business
Bitnob’s site now sells to businesses too. Fonbnk, long known for turning prepaid airtime into crypto, now calls itself “the stablecoin settlement layer for global commerce” and lists airtime as one rail among several. Kotani Pay says it is registered as a Financial Service Provider in South Africa (number 53594). Its co-founder Felix Macharia told a Nairobi conference in 2026: “What stablecoins are improving today is cross-border movement, treasury operations and settlement speed.” That sentence describes plumbing, not a replacement for the app on your phone.
Which means the ordinary sender may meet stablecoins without knowing it, inside a remittance app that settles in USDC behind the scenes. The incumbents are moving as well: Genfinity reported in May 2026 that Western Union’s chief executive had confirmed that month’s launch of USDPT, a dollar token on Solana issued by Anchorage Digital Bank, first as an internal way to settle with its agents rather than a consumer product. If a licensed firm uses a stablecoin rail and passes on the saving, the sender gets the benefit with none of the four legs to manage, and the digital-only benchmark above is where that saving will show.
How the mobile money leg really works
You cannot send USDT to an M-Pesa number. M-Pesa, MTN MoMo, Airtel Money and Orange Money wallets hold local money, so a third party has to stand between the chain and the wallet.
So when people ask how to send USDT to mobile money, there are two real answers.
- A payout company takes the tokens and credits the wallet in local currency through its own business connection to the mobile money network. For the recipient it is an ordinary credit. Fonbnk’s site lists such rails, naming M-PESA in Kenya and mobile money in markets such as Ghana, Tanzania, Senegal and Cameroon.
- A person trades with you. On a P2P market you release USDT from escrow once a stranger has sent mobile money to the recipient. Our Cameroon guide counted 121 offers to sell USDT for CFA francs on Binance’s P2P market at 09:59 UTC on 26 October 2026, most of them paid through MTN Mobile Money.
Neither route means the operator supports crypto. Operators have started to touch stablecoins, but the public moves reviewed for this lesson, as of 2 November 2026, sit behind the screen or remain plans, and none gives a customer a way to sell USDT into a wallet:
- M-Pesa signed with Abu Dhabi’s ADI Foundation in January 2026 to bring its blockchain and a dirham-backed stablecoin to eight markets for cross-border payments. Business Daily noted that M-Pesa and Safaricom had kept “a cautious approach to direct cryptocurrency integration” over regulatory concerns.
- In July 2026 Visa, M-Pesa Africa and Onafriq began a pilot in the DR Congo. Wallet top-ups are settled in stablecoins behind the scenes, and customers keep using mobile money as before.
- Onafriq, the payments network that links many African wallets, offers partners USDC as a settlement option. Circle’s account of the project speaks of institutions, not of people holding USDC.
- MTN and Airtel Africa have announced plans for stablecoins and crypto-linked services, ITWeb reported in February 2026. Airtel was reported as exploring the idea with no launch date.
For the family at the far end, then, nothing changes yet. Wallet limits, identity tiers and the operator’s own rules still apply to the money once it lands, and a payout firm’s access depends on its licences and partners, which can change. In Cameroon the point goes further, as the rules section below shows.
P2P, frozen accounts and fraud
P2P is where the cheap rate often lives, and where the worst outcomes happen, because the person cashing out depends on a stranger’s money being clean.
Nigeria supplied the first large case. On 23 April 2024 the chairman of the Economic and Financial Crimes Commission said about 300 accounts had been blocked over peer-to-peer foreign exchange trades, and a Federal High Court order of 24 April allowed the block of 1,146 accounts, according to TechCabal, which quoted a bank telling a frozen customer: “It has been confirmed that you trade cryptocurrency.” Two Binance executives had been charged that year. The money-laundering charges against Tigran Gambaryan were dropped in October 2024 and officials said at the time that they would keep pursuing a tax case against Binance.
Kenya followed in April 2026. TechCabal reported, citing senior investigators, that the Directorate of Criminal Investigations had frozen an undisclosed number of Binance accounts in a crackdown on crypto-linked fraud and money laundering, and users said funds tied up in P2P trades were out of reach. Binance told Business Daily that restrictions may follow “requests from relevant authorities” and declined to discuss the investigation. Business Daily’s report gave no specific legal basis.
The two episodes teach the same thing. A freeze can hit an innocent recipient whose counterparty’s funds were dirty, and then the family loses access to more than the remittance. Scams are the other risk, from fake payment screenshots to reversed transfers and false support agents; our guide on how to avoid crypto scams lists the common forms. A few habits lower the odds without removing them:
- Send a small first transfer and confirm the local-currency amount before sending more.
- Prefer a licensed firm where the receiving country has a licensing regime, and check the register yourself.
- Keep receipts on both sides: amounts, dates, addresses, the counterparty and the reason for the payment.
- Never share a seed phrase, and never pay a service that promises to recover frozen or stolen funds.
Lightning: bitcoin that fits a basic phone
Stablecoins are not the only option. Lightning, the payment network built on top of bitcoin, moves value for very little and settles in seconds, and one product was built for African phones. Machankura lets people send and receive sats with a phone number, by USSD in Côte d’Ivoire, Ghana, Kenya, Malawi, Namibia, Nigeria and Zambia, by SMS in Tanzania and Uganda, and through a website in South Africa. The site says users without an internet-connected device can redeem Azteco vouchers for bitcoin, which is one way to load it.
Strike tried the mainstream version. In December 2022 it launched Send Globally with Bitnob for US users sending to Nigeria, Kenya and Ghana: Lightning carried the value, and local currency reached a bank, mobile money or Bitnob account. Strike’s site now lists a launch in seven African countries in February 2024 and availability in over 100 countries, without naming African payout corridors, so confirm in the app what a given route does today.
What Lightning changes is the middle leg and the price risk, which shrinks to the seconds the payment takes when the sender converts right away. What it leaves alone is both doors. If the recipient needs shillings, somebody still has to buy that bitcoin, at a spread. And a recipient who holds sats instead of selling them carries bitcoin’s full volatility, which a family budget rarely can.
Rules on both sides of the transfer
A crypto transfer is not outside the law because it crosses a blockchain. Firms at each door carry anti-money-laundering duties under what regulators call the Travel Rule, and KYC identity checks apply. This is the state of play as of 2 November 2026, and it is general information, not legal advice.
| Where | Rule | What it means for you |
|---|---|---|
| United States | FinCEN guidance of May 2019: the Funds Travel Rule covers transmittals of $3,000 or more, including in crypto | Falls on providers acting as money transmitters. Expect ID checks and questions at the exchange |
| United States | 1% federal remittance tax (26 USC 4475) on remittance transfers made after 31 December 2025 | The sender pays and the provider collects, only when the sender hands the provider cash, a money order, a cashier’s check or a similar physical instrument. Funds from an account at a covered financial institution, or a US-issued debit or credit card, are exempt, so a stablecoin bought that way does not trigger it. Rules proposed in April 2026 would add traveler’s checks |
| United Kingdom | Travel Rule since 1 September 2023 | UK cryptoasset firms collect, verify and share transfer data, and keep it even when the other country has no such rule |
| European Union | Regulation 2023/1113, EBA guidelines from 30 December 2024; MiCA licensing | Transfers above EUR 1,000 to or from a self-hosted address trigger an ownership check. See our guide to MiCA |
| Kenya | VASP Act 2025, in force 4 November 2025; Regulations of 22 July 2026 | Licensing mandatory; existing providers had one year from commencement, a period Digital Policy Alert dates to 4 November 2026. See our page on the deadline |
| Nigeria | Investments and Securities Act, March 2025 | The SEC is the main regulator of crypto firms. A July 2026 executive order created a coordinating council chaired by the central bank |
| Ghana | VASP Act 2025 (Act 1154) | Bank of Ghana names it as the basis for licensing; its application page was still not up on 2 November 2026 |
| South Africa | Crypto assets treated as a financial product; FSCA licensing since June 2023 | 310 licences approved of 533 applications as of 31 March 2026 |
The franc zones need their own note. In Central Africa, the banking commission has since 6 May 2022 barred the banks, microfinance and payment institutions it supervises from holding, exchanging or converting crypto, and no digital asset licence had been granted as of June 2026; our page on crypto in Cameroon and the CEMAC goes through each text, and a later lesson in this chapter works through a Douala merchant’s case. In West Africa, a licence is required under a 2023 anti-money-laundering law but no procedure exists to obtain one, and the central bank warned against crypto on 22 July 2026: see our guide to crypto in Senegal and Ivory Coast.
Both CFA francs are also pegged to the euro at 655.957 francs for one euro. For a sender in Europe, a dollar token adds a conversion into dollars and another out of them on a corridor where the currencies are fixed against each other. This is our reasoning rather than a published finding, but the logic is simple: on euro-to-CFA routes the stablecoin has less to fix. Rules on moving foreign currency apply in many receiving countries, and a stablecoin does not lift them.
Tax for senders in the US and UK
Tax depends on what you do with the token, not on the chain it crosses. The IRS FAQ says the recipient of a genuine gift of virtual currency has no income until they sell or dispose of it. Selling another crypto to fund the transfer is a taxable disposal, as our guide to US crypto taxes explains. In the UK, HMRC treats giving tokens to anyone but a spouse or civil partner as a disposal, valued in pounds, though the gain on a stablecoin bought minutes earlier should be close to nothing; see our UK crypto tax guide. If you send regularly, keep the records and ask a tax professional.
When a stablecoin route makes sense, and when it does not
There is no general answer on crypto remittances to Africa. The factors below decide it for a given transfer, and none of this is financial advice.
- It can make sense if you already earn or hold stablecoins, since the first door is behind you and each conversion avoided saves a spread.
- It can make sense on a corridor with few licensed services or a high quoted cost, provided a lawful off-ramp exists at the far end.
- It rarely makes sense if you must buy with a card and sell through P2P on a route where digital-only firms already charge 3 to 4 percent.
- It makes little sense for a recipient who has never used a crypto wallet, since one wrong network or a fake support message can cost the whole amount, with no agent to call.
- It deserves a second look where the receiving country or the banks treat crypto with hostility, because a frozen account costs more than any saving.
The next development to watch is less visible than any app: licensed remittance firms settling in stablecoins and competing the saving away. Watch the end of Kenya’s transition on 4 November 2026, the Visa pilot’s results in the DR Congo and the World Bank’s next issue of Remittance Prices Worldwide, which will show whether the 8.46 percent has moved. The next lesson in this course turns from moving money to borrowing it, with credit on-chain in emerging markets.
