Lightning Network Payments: How Merchants Accept Bitcoin Now

For most of the last decade, Bitcoin’s Lightning Network has been a story about wallets, nodes and channel capacity. In 2026, it is increasingly a story about registers. A Midwestern burger chain now settles a meaningful share of its sales over Lightning rails and has started keeping some of the proceeds in a corporate treasury. A Bitcoin media company turned its own conference infrastructure into a live payments testbed and set a Guinness World Record doing it. The largest point of sale company in the United States has switched Bitcoin acceptance on for roughly a million merchants, with several million more targeted by the end of the year. And outside the United States, Lightning has become payroll infrastructure for tens of thousands of workers who never think about it as a crypto product at all.

None of this means Lightning has replaced Visa or Mastercard at checkout, and nobody serious is claiming it will anytime soon. What has changed is that the plumbing letting a customer tap a phone and settle a payment in under a second, with no chargeback and a fee measured in fractions of a cent, has moved from conference hall demo to production infrastructure at a scale worth taking seriously. This piece looks at how merchants are actually accepting Lightning payments right now: who is doing it, what it costs against the card networks, how the liquidity and custody choices work, and where stablecoins and regulation complicate the picture.

Why Card Fees Are Pushing Merchants Toward Lightning

The starting point for almost every merchant conversation about Lightning is not Bitcoin at all. It is the cost of accepting a card. Under 2026 interchange-plus pricing, the blended effective rate for Visa, Mastercard and Discover transactions runs at roughly 1.79 percent plus 8 cents for an in-person swipe or tap, and climbs to about 2.31 percent plus 25 cents for a card-not-present online sale, according to Motley Fool’s 2026 processing cost research. American Express runs higher still, at 2.59 to 2.99 percent depending on channel. The combined Visa and Mastercard average interchange rate reached 2.36 percent in 2025 by the same accounting, the single largest line item in most small retailers’ cost of doing business after labor and inventory.

For a restaurant running on a single-digit net margin, a couple of percentage points on every sale is not a rounding error. It is the difference between a profitable location and a marginal one. Card payments also carry a second, less visible cost: chargebacks, where a customer disputes a transaction after the fact and a merchant can lose both the goods and the payment. Lightning payments, like any Bitcoin transaction, settle final once confirmed, with no dispute mechanism built into the protocol for a merchant to worry about. That combination, a lower headline cost plus no chargeback exposure, is what has moved Lightning from a novelty at Bitcoin conferences to a genuine line item on payment strategy decks at chains like Steak ‘n Shake.

The Fee Math: Lightning Versus the Card Networks

The actual cost of a Lightning payment has two components. The first is the network’s own routing fee, which recent data from Lightning analytics firm Spark puts at a median base fee of well under a cent and a median fee rate around 143 parts per million, in practice a rounding error on a typical retail transaction, per Spark’s 2026 network research. The second is whatever markup a merchant’s processor adds on top for handling node operations, liquidity and any fiat conversion, and that figure varies by provider. Even with a processor markup included, the combined cost still tends to land well under the blended card rate, which is the arithmetic behind Steak ‘n Shake’s own reported savings of nearly 50 percent versus its card processing costs, according to Speed’s case study on the rollout.

That table comes with a caveat worth stating plainly. Near zero describes the network’s own routing fee, not necessarily the all-in price a merchant pays once a processor’s margin is added on top. A managed provider still needs to earn revenue for sourcing liquidity, handling compliance and keeping infrastructure online around the clock, so the honest comparison is processor markup versus card interchange, not protocol fee versus interchange. In practice that markup has consistently landed well below card economics across every deployment covered in this piece, which is why the savings claims, company reported rather than independently audited, read as broadly plausible rather than promotional exaggeration.

Case Study: Steak ‘n Shake’s Bitcoin Experiment Gets Bigger

Steak ‘n Shake, the roughly 400-location burger chain owned by Sardar Biglari’s Biglari Holdings, switched on Bitcoin payments over Lightning across its US locations on May 16, 2025, with the first transaction processed through the ZeusLN wallet. The deployment runs on infrastructure from Speed, a Tether-backed payments processor, across kiosks, counters and drive-throughs at more than 300 restaurants, letting customers pay by scanning a Lightning-enabled QR code with settlement described as instant.

The company has reported nearly 50 percent lower processing costs than its card fees and same-store sales up roughly 15 percent since the rollout, according to The Block’s coverage; those figures are company reported rather than independently audited, but they are directionally consistent with the fee math above. On October 31, 2025, the chain added a rewards layer through a partnership with Fold Holdings, giving customers $5 in Bitcoin for qualifying purchases.

The more interesting move came eight months after launch. On January 17, 2026, Biglari Holdings disclosed a $10 million purchase of roughly 105 BTC for its corporate treasury, its first publicly disclosed direct Bitcoin allocation, with a stated policy of routing customer Bitcoin payments into that Strategic Bitcoin Reserve rather than converting all of them to cash. That puts Steak ‘n Shake in the company of more than 200 public companies now holding Bitcoin on their balance sheets, though its position is modest next to the largest holders. Treasury decisions of this kind increasingly touch the same institutional trading venues used by funds managing far larger allocations, a dynamic HOGE Wire examined in its look at how Coinbase, Binance, Kraken and OKX handle institutional trading. The chain has since teased a Bitcoin Milkshake for the Bitcoin 2026 conference, treating its payments story as marketing as much as infrastructure.

Case Study: BTC Inc Turns Its Own Events Into Payment Infrastructure

If Steak ‘n Shake shows a mainstream retailer choosing a managed processor, BTC Inc, the company behind Bitcoin Magazine and the Bitcoin conference series, shows the opposite approach: run the open source, self-hosted stack yourself. On April 22, 2026, BTC Inc announced it had integrated Lightning payments into its own BTCPay Server infrastructure across conference ticketing, onsite point of sale and e-commerce ahead of Bitcoin 2026 at the Venetian in Las Vegas, the first time the company unified Lightning across all of its revenue lines on a single payment stack, according to crypto.news.

Di Lewis, BTC Inc’s CFO, said BTCPay Server was “the first infrastructure we’ve had that’s flexible enough to fit all of it,” while CEO Brandon Green called Lightning “the next layer” on top of the foundation the company has built. BTCPay Server founder Nicolas Dorier framed the appeal in developer terms: because the software is open source, “developers and users are free to extend and use BTCPay Server in ways even we did not foresee.”

The track record behind that confidence is real. BTC Inc’s VendorPay plugin processed more than $1 million in vendor and staff payouts at Bitcoin Asia 2024 in Hong Kong, and at Bitcoin Conference 2025 in Las Vegas the company set a Guinness World Record with 4,187 Lightning and NFC Bolt Card transactions processed in eight hours. Across its events overall, BTC Inc has logged more than 5,600 in-person Bitcoin transactions. Unlike Steak ‘n Shake or Block, nobody at BTC Inc is collecting a processor fee on these transactions; the tradeoff for zero platform cost is that the company runs and maintains its own Lightning infrastructure rather than outsourcing it.

Block’s Bet: Bitcoin Payments for Millions of Square Sellers

The scale play belongs to Block. Jack Dorsey announced the company’s plan to bring Bitcoin payments to Square sellers at the Bitcoin 2025 conference in Las Vegas on May 27, 2025, posting on X that “merchants can choose to hold the Bitcoin, or auto-convert it to fiat in real-time,” per The Defiant’s coverage of the announcement. CoinDesk reported the rollout began in earnest with a soft launch on October 8, 2025, followed by a wider release starting November 10, 2025, with Block waiving its usual processing fee on Bitcoin transactions into 2027 as an incentive to adopt early.

The bigger shift came in March 2026, when Block moved from opt-in enrollment to automatically enabling Bitcoin acceptance for eligible sellers starting March 30. By May 2026, that auto-enrollment push had switched on Bitcoin payments for roughly one million US sellers, according to crypto.news, on the way to a stated company target of more than four million merchants. The mechanics require no new hardware: existing Square terminals display a QR code at checkout, a customer pays over Lightning, and the merchant receives US dollars by default unless it specifically opts to hold Bitcoin, mirroring Dorsey’s original framing.

Put the three case studies side by side and a pattern emerges. BTC Inc runs its own infrastructure for maximum control and zero platform fees. Steak ‘n Shake outsources to a managed processor for speed of deployment across hundreds of locations. Block skips the integration question entirely by pushing Bitcoin acceptance out as a software update to a point of sale system merchants already use. All three lean on the same underlying rail; they differ only in who operates the plumbing.

Beyond the United States: Cross-Border Merchants and Payroll

The US retail examples above are the highest profile because they involve household names, but Lightning’s most consistent commercial use case predates all of them by years and lives outside the United States entirely. Bitnob, active across 23 African countries, has grown Lightning-based remittance and payroll volume roughly 340 percent year over year, letting employers pay remote workers and families send money home without touching the traditional remittance corridors that the World Bank’s own tracking still shows charging some of the highest fees in the world, particularly across Sub-Saharan Africa. Businesses in that context are not choosing Lightning over a card network; they are choosing it over a wire transfer or a cash pickup service that can take days and charge a double-digit percentage fee.

El Salvador remains the most watched government-level experiment, even as its state-run Chivo wallet is being wound down or sold under the terms of the country’s IMF financing program. Bitcoin stays legal tender and private wallets continue serving merchants and consumers regardless of what happens to Chivo specifically. The pattern across both examples is the same one visible in the US case studies: Lightning tends to win first wherever the alternative is genuinely bad, whether that is a steep remittance fee or a multi-percent card swipe, rather than displacing payment rails that already work reasonably well.

Choosing How to Accept Lightning: Self-Hosted Versus Managed

For a merchant deciding how to actually get started, the three models above map onto three real choices.

• Self-hosted (BTCPay Server): free, open source, non-custodial software the merchant runs itself, with no platform fee and no third party ever touching the funds, at the cost of technical setup and ongoing operational responsibility.
• Managed processor (Speed, OpenNode): an API or plugin integration where the provider handles node operations, liquidity and optional fiat conversion, trading a processing fee for a deployment Speed says it can complete in a matter of days.
• Integrated point of sale (Block / Square): Bitcoin acceptance arrives as a software update inside a point of sale system the merchant already runs, requiring no new integration work but limiting the merchant to whatever that platform supports.

Voltage occupies a middle position worth calling out separately. Rather than a checkout gateway, it functions more as managed node infrastructure, and in February 2026 it launched Voltage Credit, a revolving line against a merchant’s Lightning cash flow with no origination fee, aimed at businesses that want the balance sheet benefits of running Bitcoin volume without the operational burden of self-hosting.

Where the Liquidity Actually Comes From

Every option above quietly depends on the same unglamorous prerequisite: inbound liquidity. A Lightning channel is a two-sided balance, and to receive a payment, whoever is on the receiving end needs channel capacity already tilted in their favor, or a just-in-time channel opened on the spot. For a hobbyist running a single node, sourcing that capacity used to be one of the network’s most persistent complaints. For a merchant using Speed, OpenNode or Voltage, the problem does not disappear; it simply moves upstream to the processor, which sources capacity through wholesale marketplaces such as Amboss Magma or Lightning Labs’ own Pool and Terminal auction system rather than leaving it to chance.

That wholesale layer is now a business in its own right, with named liquidity providers, escrow mechanisms and pricing that behaves like any other short-term capital market. HOGE Wire covered that marketplace in detail in Lightning’s Liquidity Market: Who Actually Supplies It, and it is worth understanding for any merchant evaluating a processor, because the depth and reliability of a provider’s liquidity sourcing is a large part of what separates a payment that clears instantly from one that stalls at the counter during a busy lunch rush.

Stablecoins at the Register

Not every merchant wants Bitcoin price exposure, even briefly. That is where stablecoins enter the picture. Speed, among other processors, already supports USDT and USDC alongside BTC across Lightning, on-chain Bitcoin and several other networks, letting a merchant quote and settle in dollar terms without touching a card network at all. The mechanism behind USDT on Lightning specifically is Taproot Assets, a Lightning Labs protocol that lets the stablecoin ride Lightning’s payment channels while Bitcoin continues to function as the underlying routing and settlement layer.

A second approach is emerging alongside it. RGB, a client-side validation protocol with roots going back to Peter Todd’s 2014 single-use-seals concept, is the foundation for a Tether-backed relaunch of USDT on Bitcoin led by the software firm UTEXO. That rollout has been described as weeks away since early July 2026 and had still not gone live on mainnet at the time of writing, a reminder that shipping timelines in this corner of crypto tend to slip. Either way, the competitive logic is the same: Bitcoin’s Lightning rails are trying to win merchant settlement volume away from Tron, Solana and the other chains that currently dominate stablecoin transfers, a contest HOGE Wire examined at length in Lightning vs. the Stablecoin Rails: Where Bitcoin Really Stands. Notably, neither Visa’s nor Mastercard’s own on-chain stablecoin settlement pilots currently include Bitcoin or Lightning among their supported networks, which is precisely the gap Lightning’s stablecoin push is aimed at closing.

Volatility, Accounting and the Tax Question

For merchants that do hold some Bitcoin, like Steak ‘n Shake’s treasury policy, price volatility is a real operational question. Most point of sale integrations sidestep it by quoting the sale price in local currency and locking in the BTC exchange rate at the moment of payment, so the merchant is not exposed to price movement unless it deliberately chooses to be, by holding rather than converting.

The less glamorous problem is recordkeeping. Every Bitcoin payment a merchant receives is a taxable event with a cost basis set at the moment of receipt in most jurisdictions, and any later sale or conversion creates a second taxable event on top of it. That is one reason processors offering automatic fiat conversion remain popular even among merchants that are otherwise enthusiastic about Bitcoin: it collapses two tax events into one and avoids tracking a moving cost basis across thousands of small transactions. For the portion of receipts that do get converted through an exchange, merchants run into the same wallet-by-wallet reporting infrastructure that consumer traders now navigate; HOGE Wire’s comparison of how Coinbase, Binance, Kraken and OKX handle tax paperwork is a useful reference for what that process actually looks like once a business is dealing with an exchange rather than a wallet it fully controls.

The Regulatory Backdrop Merchants Cannot Ignore

Payment stablecoins, the dollar-settlement option described above, now sit inside the framework created by the GENIUS Act, signed into law on July 18, 2025 (S. 1582, Congress.gov). The law gave six primary regulators, including the OCC, FDIC and Treasury’s FinCEN and OFAC arms, one year to finalize implementing rules. That deadline passed on July 18, 2026 with none of the rules final; several agencies had only published additional proposed rules with comment windows stretching past the deadline itself. The law’s substantive requirements still activate automatically by January 18, 2027, or 120 days after final rules are published, whichever comes first, so the rulebook is delayed rather than optional.

For a merchant, the practical takeaway is narrower than it sounds. Accepting and holding Bitcoin itself is not governed by any of this, since Bitcoin is not a stablecoin and carries no issuer. It matters only if a merchant or its processor wants to settle specifically in USDT or another payment stablecoin, where Tether’s status as a foreign issuer keeps its long-term US access an open compliance question. Separately, any processor that touches fiat conversion is a money services business subject to the same state money-transmission licensing and FinCEN registration payment companies have navigated for years, a compliance layer that predates crypto and is not unique to it. Merchants operating in the European Union face a different but parallel structure: MiCA governs stablecoin issuers and crypto-asset service providers rather than the merchant itself, so a payment processor operating there needs CASP registration much like an exchange would, while the shop accepting the payment generally does not.

Security and Custody: What Can Still Go Wrong

The self-hosted and managed paths carry genuinely different risk profiles. Running BTCPay Server keeps private keys with the merchant, which removes counterparty risk from a processor but places full responsibility for backups, channel management and uptime on a business whose core competency is usually not running payments infrastructure. A managed processor removes that operational burden but reintroduces a form of counterparty reliance: funds pass through the provider before reaching a merchant’s bank account or wallet, which makes a provider’s licensing, track record and insurance arrangements worth real diligence before signing up.

At the protocol level, Lightning’s known weak points remain unsolved in the academic sense but unexploited at meaningful scale according to public reporting so far. Channel jamming is the clearest example: an attacker routes a payment to itself and deliberately withholds settlement, tying up a channel’s limited payment slots (a hard cap of 483 pending payments per direction, inherited from a limit on Bitcoin’s own transaction size) until a timelock expires. Researchers have proposed reputation and bonding systems to raise the cost of the attack, but no fix is universally deployed yet. For a merchant, the practical exposure is indirect: a jammed channel somewhere upstream can slow or fail a payment, which is one more reason a processor’s liquidity depth and route diversity matter more than any single technical safeguard. The more mundane risk sits at the actual counter: fake QR codes, clipboard-hijacking malware that swaps a copied Lightning invoice for an attacker’s own, and basic staff training gaps are a far more likely source of loss for a retail deployment than a protocol-level exploit.

What Is Still Missing Before Lightning Goes Mainstream

The case studies above are genuinely impressive relative to where Lightning stood even two years ago, but a few gaps keep merchant acceptance from translating into everyday consumer use.

• Wallet adoption: most shoppers still do not carry a funded Lightning wallet ready to spend, so merchant acceptance alone does not create demand on its own.
• Fragmentation: BTCPay Server, Speed, OpenNode, Voltage and Block’s own integration each behave slightly differently at checkout, so there is no single unified merchant experience yet.
• Scale: despite Block’s push, no major national grocery chain or big box retailer has adopted at anything close to Steak ‘n Shake’s depth of integration.
• Perception: volatility concerns persist among both merchants and shoppers even where the operational exposure has effectively been solved by auto-conversion.
• Regulatory caution: the unresolved stablecoin rulebook described above keeps some processors conservative about which assets they support and where.

What Comes Next

The next twelve months carry a few concrete markers to watch. Block’s push toward its four million merchant target will be the clearest signal of whether mainstream point of sale adoption is real or a rounding error against the total US small business count. UTEXO’s RGB-based USDT rollout, if it finally ships after a month of weeks-away timelines, would give merchants a second dollar-settlement rail alongside Taproot Assets and add real competition to that side of the stack.

Further out, the same rails now handling a human tapping a phone at a Steak ‘n Shake counter are being extended toward payments that do not involve a human at all. Lightning Labs’ own Wavelength protocol, alongside the broader L402 agent-payment ecosystem, is explicitly aimed at letting AI agents and connected devices make sub-cent payments without a person approving each one, a use case HOGE Wire covered in Lightning Labs’ Wavelength: Bitcoin’s Third Ark for AI Agents. The throughline across all of it is that Lightning’s merchant story in 2026 is no longer about proving the technology works. It is about distribution, and distribution is a slower, less glamorous fight than any protocol upgrade.

Frequently Asked Questions

What is the Lightning Network and how does it help merchants accept Bitcoin?

Lightning is a payment layer built on top of Bitcoin that lets participants transact through payment channels instead of the base blockchain, settling in under a second with fees typically far below a cent. That speed and cost make it usable for point of sale purchases, where paying directly on Bitcoin’s base layer would be both too slow and too expensive per transaction for everyday retail. Merchants connect through either self-hosted software like BTCPay Server or a managed processor, and customers pay by scanning a QR code with any Lightning-enabled wallet.

How much does it cost a merchant to accept Bitcoin over Lightning?

Routing fees on the Lightning Network itself are typically a small fraction of a percent of the transaction value, with recent network data putting the median fee rate around 143 parts per million. Most merchants also pay an additional fee to whichever processor they use, and that markup varies by provider. Even with a processor fee included, businesses like Steak ‘n Shake have reported costs near half of what they pay for card processing, though exact savings depend on the processor chosen and the merchant’s overall payment mix.

Do merchants have to hold Bitcoin, or can they get paid in dollars?

Most processors, including Block’s Square integration and gateways like OpenNode and Speed, let a merchant automatically convert every Bitcoin payment to fiat currency in real time. Accepting Lightning payments does not require taking on Bitcoin price risk unless a merchant deliberately chooses to hold some or all of what it receives, as Steak ‘n Shake has done with part of its revenue since establishing a Strategic Bitcoin Reserve in January 2026.

What is the difference between BTCPay Server and a processor like Speed or OpenNode?

BTCPay Server is free, open source, self-hosted software that lets a merchant run its own non-custodial Lightning node, with no platform fee and no third party ever holding the funds. Managed processors such as Speed and OpenNode instead handle node operations, liquidity and settlement on the merchant’s behalf through an API or plugin, which is faster to deploy but means funds briefly pass through the provider before reaching the merchant.

Can customers pay a Lightning merchant with USDT instead of Bitcoin?

Increasingly, yes. Taproot Assets, a Lightning Labs protocol, lets Tether’s USDT move over Lightning’s payment channels while Bitcoin remains the underlying settlement layer, and some processors already support USDT alongside BTC. A second Bitcoin-based dollar rail built on the RGB protocol has been described as close to a commercial launch for several weeks, though it was not yet live on every processor at the time of writing.

Written by the HOGE Wire crypto desk.