04
Bitcoin for a Cup of Coffee
The remittance worker we met in Section 1, Joanne, sending $500 a month from San Francisco to Cebu, doesn't actually settle her transfers on Bitcoin's base layer. She uses Lightning. The settlement clears in under five seconds, costs less than a US cent, and never touches a bank, yet it ends with Filipino pesos in her parents' phone wallet. That isn't a feature of Bitcoin per se. It is a feature of a payment network that lives on top of Bitcoin called the Lightning Network, and it is the answer to a question Bitcoin's base layer is, by design, the wrong tool for.
Bitcoin's base layer is intentionally slow. It settles a new block every ten minutes, with a global throughput ceiling of roughly seven transactions per second, and that ceiling is deliberate. Small, infrequent blocks are easy for individual node operators to verify, which is what keeps the network decentralized. But seven transactions per second is enough for the world's savings to settle. It is not enough for the world's spending. If everyone tried to buy coffee on the base layer, most of those coffees would still be waiting in line by closing time.
Lightning solves that by moving the small transactions off-chain. Two parties open a payment channel, a two-person ledger backed by a single on-chain Bitcoin transaction. Inside the channel they can send each other bitcoin instantly, in any amount, for a fraction of a cent per transaction, as many times as they want. Final balances settle to the main blockchain only when the channel closes. Now connect roughly 14,940 of those nodes1 through 48,678 channels, and the channels can route payments across the network the way packets route across the internet, hopping node to node until the payment reaches its destination. The end user doesn't see any of that. They scan a QR code, the money arrives, the balance updates.
5,100
BTC of public Lightning capacity
14,940
Public Lightning nodes
48,678
Public Lightning channels
Static fallbacks shown as of April 2026; the HTML version updates these from mempool.space on page load. ATH public capacity reached 5,637 BTC in 2025.
Public capacity is a floor, not a ceiling. Strike, Cash App, Kraken, Bitfinex, and Binance all run private routing capacity that doesn't show up on the public dashboards, and that capacity is estimated to be multiples of the public number2. Lightning's actual reach is significantly larger than what the public-node statistics make it look.
Soweto, 2023: Machankura
In 2023, a South African software engineer named Kgothatso Ngako built a service called Machankura3. Machankura lets anyone with a basic feature phone (the kind that costs $20 used, runs on 2G, has no smartphone capabilities, no internet, and no app store) receive and send bitcoin on the Lightning Network using nothing but USSD codes (the *#-style menus that have run mobile banking and airtime top-ups in Africa for two decades).
By 2025 the service was operating across eight African countries, including South Africa, Kenya, Nigeria, Ghana, Uganda, Zambia, Malawi, and Tanzania. The user types a short menu code, follows a few prompts, and is now part of the Lightning Network, able to receive bitcoin from anywhere in the world, in seconds, with no smartphone, no Wi-Fi, and no bank.
This is the part of the Bitcoin story that the “developing-world adoption” framing gets backwards. Machankura is not adoption-with-asterisks. It is bitcoin running on 2005-era hardware, in regions where ~80% of adults own a feature phone and a large share have no bank account, and it works.
Lightning is the final missing layer. Once you can pay a penny in a second to a stranger on the other side of the world, money becomes what it always should have been.
Elizabeth Stark, CEO, Lightning Labs
Two layers, one money. Bitcoin's base layer is the global settlement system, analogous to the Federal Reserve's wire transfer network in the United States. It is slow, expensive per transaction relative to its volume, and used to settle large or final balances. Lightning is the everyday spending layer, analogous to a card swipe. It is fast, cheap, and routes through a payments network. The difference is that there is no Visa, no Mastercard, no bank, and no permission-granting institution between you and your counterparty. The settlement layer and the spending layer are both the same money. They are not two different systems pretending to interoperate.
The Landing
Lightning is not futuristic. As you read this, it is hiding inside the most under-banked parts of the world: on feature phones in Soweto, in remittance apps in Manila, in coffee shops in El Zonte, in payouts on Strike, in cross-border invoices that no bank ever saw. Bitcoin's base layer is what makes Lightning possible. Lightning is what makes Bitcoin practical.
¹ mempool.space Lightning dashboard, April 2026 snapshot, mempool.space/lightning; cross-referenced with 1ML.com Lightning Network statistics, 1ml.com. Public capacity all-time high: 5,637 BTC, reached in 2025.
² Public-vs-private capacity estimate: industry reporting from Lightning Labs, Voltage, and Amboss; large custodial routing capacity (Strike, Cash App, Kraken, Bitfinex, Binance) is not reflected in public dashboards.
³ Machankura, service homepage 8333.mobi; profile coverage in Bitcoin Magazine, CoinDesk, and Decrypt 2023–2025. Founder: Kgothatso Ngako.