How to Measure a Payment Network That Values Privacy
Tracking the Lightning Network is fundamentally different from tracking traditional payment systems. Visa publishes quarterly transaction volumes. SWIFT reports message counts. The Lightning Network, by design, makes comprehensive measurement difficult. Payments travel through encrypted onion-routed paths. Private channels do not appear in the public graph. Transaction amounts and sender-receiver pairs are invisible to outside observers.
Every metric discussed here comes with a caveat: public data captures only a fraction of actual network activity. Private channels account for a substantial portion of the network's real capacity and transaction volume. The numbers reported here represent the visible floor, not the ceiling.
The primary metrics available for public analysis include:
- Network capacity: Total bitcoin locked in public payment channels
- Public channels: The number of open, publicly announced payment channels
- Public nodes: The number of Lightning nodes visible on the network graph, discovered through the gossip protocol defined in BOLT 7
- Transaction volume: Estimated from routing node operators, payment processors, and exchange data
Network Capacity: Growth, Peak, and Consolidation
Lightning Network capacity has followed a path familiar to many maturing networks: rapid early growth, a peak, and a multi-year period of consolidation below that peak. Public capacity has stabilized in a band well short of its all-time high, even as the underlying infrastructure has matured.
Historical Growth
When the Lightning Network launched on Bitcoin mainnet in early 2018, capacity was measured in single-digit bitcoin. By the end of 2020, it had crossed 1,000 BTC. The bull market of 2021, amplified by El Salvador's adoption of Bitcoin as legal tender (legal tender status was repealed in January 2025 as a condition of IMF lending; merchant adoption continues), pushed capacity past 3,300 BTC. It peaked above 5,400 BTC in early 2023 before entering a multi-year correction.
The 2025 Capacity Dip and Late-2025 Spike
Through the first half of 2025, capacity declined approximately 20%, falling to around 4,200 BTC by mid-year. This drew attention from observers who interpreted the decline as a sign of network weakness.
Several factors contributed to the temporary decline:
- Channel consolidation: Operators closed smaller, underutilized channels in favor of fewer, larger, better-capitalized ones. The average channel size increased even as channel count fell.
- Routing efficiency improvements: Protocol upgrades and better pathfinding algorithms allowed the same transaction volume to flow through fewer channels.
- Bitcoin price appreciation: As the bitcoin price rose, the dollar-denominated capacity of the network increased even when BTC-denominated capacity was flat or declining.
- Splicing adoption: Channel splicing, which allows adding or removing funds from a channel without closing it, reduced the need for redundant channel openings.
In late 2025, capacity briefly rebounded to a new all-time high of approximately 5,637 BTC, driven largely by major exchanges (including Binance and OKX) adding liquidity to the network. That spike proved temporary. By mid-2026, public capacity had declined again to approximately 4,800 BTC (mempool.space), in line with the range it has occupied since the 2023 peak. The honest reading is that public capacity has consolidated rather than grown steadily: it remains below the levels reached at the 2022-2023 peak.
Under the Hood: Announced vs. Unannounced Channels
The Lightning Network distinguishes between announced (public) and unannounced (private) channels. When two nodes open a channel, they can choose whether to broadcast a channel_announcement message via the BOLT 7 gossip protocol. Announced channels appear in every node's channel graph and are available for routing third-party payments. Unannounced channels are known only to the two participants and are not used for routing other people's payments.
Public capacity metrics capture only announced channels. Research estimates suggest that unannounced channels account for a substantial portion of total network capacity, potentially 30-50% or more. Mobile wallets using Lightning Service Providers (LSPs) typically use unannounced channels, since these channels serve only the wallet holder and do not need to participate in the public routing network. This means the visible network is a floor estimate, not a complete picture.
Wumbo channels (channels exceeding the original 0.1677 BTC default limit, enabled by the option_support_large_channel feature bit) have also shifted the capacity distribution. Before Wumbo support was widespread, channels were capped at approximately 16.7 million satoshis. Today, channels of 1 BTC or more are common among professional routing nodes, concentrating capacity in fewer, larger channels.
What Capacity Means (and What It Does Not)
Capacity is the total bitcoin locked in public channels, which sets an upper bound on how much those channels can carry at once. Within any single channel the funds sit on one side or the other, so the amount actually sendable in a given direction is always less than the channel's full capacity. It is not a direct measure of how much bitcoin moves through the network over time. A 1 BTC channel can route hundreds of bitcoin per month if payments flow in both directions and rebalance naturally.
Think of capacity like the width of a highway. A two-lane road can carry far more total traffic over a day than the number of cars it holds at any single moment, because cars enter and exit continuously. Lightning capacity works the same way: the flow of payments through channels over time far exceeds the static balance locked in them.
Nodes and Channels: Consolidation, Not Decline
Public node count peaked around 20,000 in early 2022 and has since settled near 17,000. Public channel count has followed a similar trajectory, dropping from over 80,000 at its peak to roughly 41,000 by mid-2026 (mempool.space).
Taken alone, these numbers might look concerning. In context, they reflect the same kind of maturation seen in every network industry. Early adopters run small, experimental nodes. As the network matures, professional routing operators replace hobbyists, running fewer but larger and better-connected nodes. The network's Gini coefficient for capacity distribution has risen to approximately 0.97, indicating that a small number of well-capitalized hubs route the majority of payments.
This pattern mirrors the evolution of internet infrastructure. Thousands of early ISPs consolidated into large backbone providers, and the internet became more efficient rather than smaller. The same dynamic is playing out on Lightning.
Adoption Milestones
El Salvador: The Lightning Proving Ground
On September 7, 2021, El Salvador became the first country to adopt Bitcoin as legal tender (legal tender status was repealed in January 2025 as a condition of IMF lending; merchant adoption continues). The government launched the Chivo wallet, a Lightning-enabled app, and distributed $30 in bitcoin to every citizen who downloaded it. Within months, three million Salvadorans (roughly 46% of the population) had downloaded the wallet.
Adoption, however, was uneven. Studies from the National Bureau of Economic Research found that while downloads were high, sustained usage was lower: approximately 37% of Chivo users made transactions on a regular basis. About 20% of businesses accepted bitcoin by 2022, and only 1.3% of remittances in 2023 flowed through Bitcoin rails.
El Salvador revealed both the potential and the friction points of Lightning-based payments at national scale. The technology worked and payments were fast and low-cost, but user education, merchant tooling, and the inertia of existing payment habits proved to be the harder problems. Every Lightning development since has been shaped, at least partly, by those lessons.
Major Platform Integrations
The more durable adoption story has come from major financial platforms integrating Lightning into their existing products:
- Cash App added Lightning send and receive in 2022 and reported substantial growth in Lightning payment volume through 2024. By mid-2025, the company reported that roughly one in four Bitcoin payments on Cash App used Lightning.
- Coinbase launched Lightning support in April 2024, powered by Lightspark. By mid-2025, Coinbase reported that around 15% of its Bitcoin transactions moved over Lightning rails.
- Strike has positioned Lightning as its core infrastructure for global payments, expanding from consumer transfers into bitcoin-backed lending and broader payment products.
These integrations matter because they bring Lightning to users who never encounter payment channels, liquidity, or routing. Users see faster, cheaper bitcoin transfers. The infrastructure underneath is Lightning.
Payment Processors and Merchant Tools
On the merchant side, payment processors have steadily expanded Lightning support:
- BTCPay Server: The open-source, self-hosted payment processor supports Lightning natively, giving merchants full control over their payment infrastructure without third-party fees.
- OpenNode: Provides Lightning payment processing for businesses, with API integrations for e-commerce platforms.
- CoinGate: Reports that the share of Bitcoin payments made through Lightning grew from 5.98% in 2022 to 7.95% in 2023 to 14.51% in 2024, a clear upward trend.
Payment Volume: What the Data Shows
Measuring Lightning payment volume is inherently imprecise because the protocol is designed to protect payment privacy. The data that does exist comes from routing node operators, payment processors, and exchanges willing to share aggregate statistics.
Available Data Points
The Lightning Network facilitated over eight million monthly transactions in early 2025. Public Lightning volume surged 266% year-over-year between 2023 and 2024, according to data aggregated by Voltage. LQWD Technologies, a publicly traded Lightning infrastructure company, reported a 61% increase in monthly routing volume through its nodes in the second half of 2024.
Block, Inc. (the parent company of Cash App and Square) reported that Lightning routing through its infrastructure yields approximately 10% annually on deployed capital, a figure that signals meaningful payment flow through the network.
The Measurement Challenge
The volume figures above represent only the transactions visible to the entities reporting them. A payment routed between two private nodes through a direct channel generates no public data at all. Multi-path payments split across several routes are counted differently by different observers. And the growing use of private channels, trampoline routing, and blinded paths (introduced with BOLT 12) further obscures the true volume.
Lightning payment volume is growing rapidly by every available metric, though the exact magnitude is unknowable by design because of the network's privacy model.
Infrastructure Growth
Lightning Implementations
Four major implementations power the Lightning Network, each targeting different use cases and operator profiles:
| Implementation | Developer | Language | Primary Use Case |
|---|---|---|---|
| LND | Lightning Labs | Go | General-purpose nodes, developer ecosystem. Runs the majority of public nodes. |
| Core Lightning (CLN) | Blockstream | C | Plugin architecture for businesses and custom integrations. Extensible and modular. |
| Eclair | ACINQ | Scala | Enterprise scalability. Powers the Phoenix wallet for mobile users. |
| LDK | Spiral (Block) | Rust | Embeddable toolkit for building custom Lightning integrations. Used by Cash App. |
Blockstream maintains Core Lightning (CLN), one of the four major Lightning implementations. CLN's plugin architecture makes it a popular choice for businesses building on Lightning, allowing operators to extend node functionality without modifying the core software. CLN led the way on BOLT 12 offers, with a full rollout in 2024 marking the first new BOLT addition since 2017.
Implementation diversity strengthens the network: each codebase independently validates the Lightning protocol specification. If a critical bug is found in one implementation, the others continue operating, mirroring Bitcoin's own decentralization philosophy.
The Wallet Ecosystem
Lightning wallet availability has expanded from a handful of technical tools to dozens of consumer-ready options:
- Self-custodial wallets: Phoenix (ACINQ), Breez, Zeus, Mutiny (now deprecated), and the Blockstream app all give users direct control of their Lightning funds.
- Custodial wallets: Wallet of Satoshi and others provide simpler onboarding at the cost of requiring trust in the wallet operator.
- Integrated wallets: Cash App, Strike, and Coinbase handle Lightning behind the scenes within broader financial apps.
Consumer adoption is supported by wallets like the Blockstream app, which provides Lightning access alongside on-chain Bitcoin and Liquid from a single interface. Cross-layer functionality lets users hold and move bitcoin across layers without switching between multiple apps.
Lightning Service Providers (LSPs)
LSPs solve one of Lightning's original friction points: the need for new users to have inbound liquidity before they can receive payments. An LSP opens channels to new users on demand, abstracting away the mechanics of channel management.
Blockstream's Greenlight, built on Core Lightning, managed over 200,000 nodes by the end of 2024, a 20x increase in a single year. Greenlight's architecture separates node execution (handled by Blockstream's infrastructure) from key management (retained on the user's device), so users get the convenience of cloud hosting with the security of self-custody. Lightspark provides production-grade Lightning infrastructure to companies like Coinbase and is working with banks in Brazil and India. Breez SDK allows any app developer to embed Lightning payments without running a node.
The LSP model mirrors how ISPs made TCP/IP accessible to consumers. The protocol complexity still exists, but a service layer absorbs it so end users never encounter it.
Institutional Interest in Lightning
Financial Services Integration
The entry of major financial institutions into Lightning infrastructure marks a shift from experimental curiosity to production deployment:
- Lightspark, founded by former PayPal president David Marcus, raised $175 million and now provides Lightning payment infrastructure to banks and fintechs globally. Partners include SoFi, Coinbase, and NuBank. Stripe and Revolut have tested Lightspark's tools.
- Banks in Brazil and India announced plans to use Lightspark for cross-border money movement in early 2025.
- Block, Inc. runs major Lightning routing infrastructure and reports commercially viable routing yields.
The Cross-Border Remittance Opportunity
The remittance market processes over $800 billion annually. Traditional corridors charge 6-10% in combined fees through SWIFT messaging, intermediary banks, and batch settlement layers. Delivery takes two to five business days. The system operates only during banking hours.
Lightning operates 24/7, settles in seconds, and charges a median fee rate of approximately 0.0029%. For a $200 remittance (the global average), traditional fees consume $12-20. Lightning fees on the same transfer amount to fractions of a cent.
| Metric | SWIFT / Traditional | Lightning Network |
|---|---|---|
| Settlement time | 2-5 business days | Seconds |
| Availability | Banking hours (varies by corridor) | 24/7/365 |
| Fee on $200 transfer | $12-20 (6-10%) | < $0.01 |
| Settlement finality | Provisional (can be reversed) | Atomic (final on completion) |
| Intermediaries required | 2-5 correspondent banks | No custodial intermediaries (routing nodes never hold funds) |
These economics explain why institutions are investing in Lightning infrastructure rather than treating it as a niche payment experiment.
Micropayments and Machine-to-Machine Payments
Traditional payment networks have a floor. Visa's interchange fees make transactions below $1-2 nonviable for merchants. Lightning has no meaningful floor. Payments of a single satoshi (a fraction of a cent) are technically possible, opening payment models that do not exist on traditional rails:
- Pay-per-article or pay-per-second streaming: Content creators can charge fractions of a cent per page view or per second of audio/video playback.
- Machine-to-machine payments: IoT devices can settle micro-charges autonomously without human intervention or credit card infrastructure.
- API metering: Developers can pay per API call in real time, replacing monthly billing cycles with usage-based micropayments.
Lightning vs. Traditional Payment Networks
Visa and SWIFT serve different functions, and Lightning's architecture differs from both.
Lightning vs. Visa
| Metric | Visa | Lightning Network |
|---|---|---|
| Transaction throughput | ~65,000 TPS capacity; ~10,000 TPS average | Theoretically unlimited (bounded by channel liquidity, not protocol) |
| Settlement finality | Authorization in seconds; final settlement in 1-3 days | Final settlement in seconds |
| Merchant fees | 1.5-3.5% of transaction value | Typically < 0.01% |
| Chargebacks | Yes (consumer protection feature and merchant cost) | No (payments are final; disputes handled at application layer) |
| Global reach | 200+ countries via merchant acquiring networks | Any internet connection; no merchant account required |
| Minimum viable payment | ~$1-2 (below this, fees exceed merchant margin) | 1 satoshi (fractions of a cent) |
The key distinction is settlement finality. When a Visa transaction is "approved" at the point of sale, the merchant has received a promise of payment rather than actual funds, and final settlement occurs days later through batch processing. A Lightning payment either fully settles or it never happened, with no intermediate state and no counterparty risk between authorization and settlement.
Lightning vs. SWIFT
SWIFT is a messaging network, not a settlement network. It transmits payment instructions between banks. The actual movement of funds happens through correspondent banking relationships that can involve multiple intermediaries, each adding time and cost.
Lightning combines messaging and settlement into a single atomic operation. A payment instruction and its settlement are the same event: the cryptographic update of channel balances between peers. This eliminates the time gap, the intermediary fees, and the counterparty risk inherent in the multi-hop correspondent banking model.
Where SWIFT requires bilateral relationships between banks and pre-funded accounts that each bank holds with the other, Lightning routes payments algorithmically through the most efficient available path with no custodial intermediaries in the loop.
Challenges to Further Adoption
Several friction points remain between Lightning's current state and mainstream payment infrastructure.
Liquidity Management
Running a Lightning node profitably requires active management of channel balances. If a channel's liquidity shifts entirely to one side, it cannot route payments in the other direction until it is rebalanced. LSPs and modern wallets handle this for individual users, but for routing node operators, liquidity management remains a specialized skill.
Onboarding Friction
Opening a Lightning channel requires an on-chain Bitcoin transaction, which means paying an on-chain fee and waiting for confirmation. During periods of high Bitcoin network fees, channel opens become expensive. Inbound liquidity (the ability to receive payments) requires either a counterparty willing to allocate funds toward the receiving node or an LSP that provides it as a service.
For experienced users, these are manageable trade-offs. For a first-time user trying to receive a $5 payment, they can be deal-breakers. Modern wallets have made progress hiding this complexity, but the underlying economics remain a constraint.
Interoperability
While the four major implementations follow the same BOLT specification, edge cases and feature adoption timelines differ. Splicing interoperability between CLN and Eclair was successfully tested in 2025, and efforts continue to ensure all implementations converge on new features like BOLT 12 offers and Simple Taproot Channels.
Capacity Concentration
The network's Gini coefficient of ~0.97 for capacity distribution means a small number of large nodes control most of the network's liquidity and routing. This creates efficiency but also introduces centralization concerns. If a small number of major routing hubs went offline simultaneously, a significant portion of the network's payment capacity would be affected. Ongoing work on multi-path payments, trampoline routing, and decentralized pathfinding algorithms aims to reduce this concentration over time.
The Road Ahead: Protocol Developments
BOLT 12 Offers
BOLT 12 replaces the static Lightning invoice (a one-time-use payment request with an expiration) with reusable offers. A merchant can publish a single offer that generates a fresh invoice for each customer automatically. BOLT 12 also introduces blinded paths, which hide the recipient's node identity from the sender, improving privacy for merchants and individuals alike.
Core Lightning's full BOLT 12 rollout in 2024 was the first new BOLT specification addition since 2017. Adoption across other implementations is underway.
Channel Splicing
Splicing allows users to add or remove funds from an existing payment channel without closing and reopening it. Before splicing, resizing a channel required two on-chain transactions (close the old channel, open a new one). With splicing, a single on-chain transaction adjusts the channel's capacity while it remains operational.
By late 2025, splicing had shipped in Core Lightning and Eclair with tested interoperability between them, while other major implementations were still adding support. Fidelity Digital Assets published research that named splicing as a key enabler for more efficient capital deployment on Lightning.
Simple Taproot Channels
Taproot channels use Bitcoin's Taproot upgrade to make Lightning channel opens and cooperative closes indistinguishable from regular Bitcoin transactions on-chain. This improves privacy for Lightning users by eliminating the on-chain fingerprint that currently identifies a transaction as a Lightning channel operation.
Channel Factories and Ark
Channel factories allow multiple participants to share a single on-chain UTXO and create bilateral Lightning channels from it, reducing the number of on-chain transactions required to establish channels. Ark, developed by Burak Keceli, takes a related approach using shared UTXOs to enable fast, cheap, off-chain bitcoin transfers that are natively interoperable with Lightning.
Both designs address the same constraint: every Lightning channel currently requires its own on-chain transaction to open. Channel factories and Ark-style protocols reduce this per-user on-chain cost, critical for onboarding the next order of magnitude of Lightning users.
Stablecoins on Lightning
Tether's January 2025 announcement of USDT on Lightning via the Taproot Assets protocol adds another dimension to Lightning's payment utility. Stablecoin payments over Lightning combine dollar-denominated stability with Lightning's speed and low fees, potentially accelerating adoption in regions where bitcoin's price volatility has been a barrier to daily use.
Frequently Asked Questions
How much bitcoin is locked in the Lightning Network?
As of mid-2026, the Lightning Network's public capacity is approximately 4,800 BTC, according to mempool.space. It briefly reached an all-time high of around 5,637 BTC in late 2025, driven largely by major exchanges adding liquidity, before declining again. This figure counts only announced channels broadcast via the BOLT 7 gossip protocol. Unannounced (private) channels, which are common among mobile wallets and LSP-connected nodes, fall outside this count. Research estimates suggest unannounced channels may account for 30-50% of additional capacity beyond the public figure. The number fluctuates as channels are opened and closed.
How many transactions does the Lightning Network process?
Public data indicates over eight million monthly transactions in early 2025, with 266% year-over-year volume growth between 2023 and 2024. Due to Lightning's privacy design, the true transaction count is higher than any single observer can measure.
Is the Lightning Network losing nodes and channels?
Public node and channel counts have declined from their 2022 peaks, but this reflects consolidation rather than collapse. Fewer, larger, better-capitalized nodes now handle more traffic more efficiently. Average capacity per channel has increased, and Wumbo channels (large channels exceeding the original 0.1677 BTC cap) have become standard among professional operators. Total network capacity briefly set an all-time high in late 2025 before settling near 4,800 BTC by mid-2026. The growth in unannounced channels used by LSP-backed mobile wallets also means that many new nodes and channels do not appear in public counts.
Which companies have integrated Lightning?
Major integrations include Coinbase (April 2024), Cash App (2022, with strong Lightning volume growth through 2024), Strike, and payment processors like BTCPay Server, OpenNode, and CoinGate. Lightspark provides Lightning infrastructure to banks and fintechs including SoFi, NuBank, and others.
How do Lightning fees compare to credit card fees?
Lightning's median fee rate is approximately 0.0029% of the transaction value, roughly 1,000 times cheaper than typical credit card interchange fees of 1.5-3.5%. For a $100 payment, a credit card charges $1.50-$3.50 in merchant fees. Lightning charges fractions of a cent.
What are the main barriers to Lightning adoption?
The primary challenges are liquidity management complexity (keeping channels balanced), onboarding friction (channel opens require on-chain fees), and the need for inbound liquidity to receive payments. LSPs and modern wallets are progressively abstracting these issues away from end users. Protocol upgrades like channel splicing (resizing channels without closing them), BOLT 12 offers (reusable payment requests), and Simple Taproot Channels (making channel operations indistinguishable from regular Bitcoin transactions on-chain) are reducing these friction points at the protocol level. For consumers, wallets like the Blockstream app remove channel management entirely, settling Lightning payments through atomic swaps so there are no channels for the user to manage, and for businesses, Blockstream's Greenlight provides managed Core Lightning infrastructure with self-custodial key control.