The Lightning Network is a Bitcoin layer-2 payment protocol that routes transactions through bidirectional payment channels off-chain, settling in milliseconds at fractions of a cent before anchoring the final balance on the Bitcoin blockchain. It is also known as BTC Lightning or LN.
Key Takeaways
- The Lightning Network operates as a routed mesh of bidirectional payment channels built on top of Bitcoin and related crypto infrastructure, keeping everyday payments off the base chain.
- As of April 2026, the public network spans 5,923 observable nodes, 18,807 open channels, and 2,741.59 BTC of aggregate capacity, according to 1ML’s network dashboard.
- The median public-channel fee rate sits near 0.000129 sat per sat transferred, which is effectively a fraction of a cent for the average retail payment.
- Joseph Poon and Thaddeus Dryja first proposed the protocol in the Lightning Network whitepaper (earliest draft February 2015, v0.5.9.2 January 2016), and Lightning Labs launched mainnet in 2018.
- River Financial’s 2023 analysis estimated a lower bound of 6.6 million routed Lightning transactions in August 2023, a 1,212% increase over the August 2021 K33 benchmark of 503,000.
How Does the Lightning Network Work?
The Lightning Network lets two parties open a shared Bitcoin funding transaction, then update the balance between them an unlimited number of times off-chain, and finally settle the last agreed balance back to Bitcoin whenever they want. Think of a payment channel as a running tab at a bar: individual rounds are tracked privately between the two of you, and only the final settlement hits the official ledger.
The Lightning whitepaper proposes a decentralized system whereby transactions are sent over a network of micropayment channels whose transfer of value occurs off-blockchain. The routing mesh, not the base blockchain, carries the everyday activity.
1. Opening a Payment Channel
Two parties commit Bitcoin to a 2-of-2 multisignature funding transaction on-chain. That single on-chain event anchors the channel. Once the funding transaction confirms, both sides hold signed commitment transactions that reflect the current channel balance and can be broadcast at any time to close the channel.
2. Sending Off-Chain Payments
Each payment updates the channel’s tentative balance by invalidating the previous commitment and signing a new one. If both parties are cooperative, the channel can remain open indefinitely, possibly for many years. Neither side publishes anything to the blockchain during normal operation, so fees stay tiny and settlement feels instant.
3. Routing via HTLCs
When the sender and receiver do not share a direct channel, the payment hops across multiple channels via Hashed Timelock Contracts. The purpose of an HTLC is to allow for a global state across multiple nodes via hashes, ensured by time commitments and time-based unencumbering of resources via disclosure of preimages. In plain English: each hop is locked to a cryptographic secret and a deadline, so the payment either clears end-to-end or unwinds safely.
4. Closing the Channel
Either party can close the channel at any time by broadcasting the latest commitment transaction. If the counterparty tries to cheat by broadcasting an older, more favorable state, a revocable sequence maturity contract lets the honest party claim the entire channel balance via a breach remedy transaction. On-chain arbitration is the ultimate backstop.
| Metric | Value (April 2026) | Source |
| Public nodes | 5,923 | 1ML |
| Public channels | 18,807 | 1ML |
| Network capacity | 2,741.59 BTC | 1ML |
| Average channels per node | 6.35 | 1ML |
| Average channel capacity | 0.145 BTC | 1ML |
| Median fee rate | 0.000129 sat/sat | 1ML |
Source: 1ML Network Statistics
Channels work because the protocol polices cheating with cryptography. The question is why that design matters for Bitcoin at all.
Why Does the Lightning Network Matter?
Bitcoin’s base chain was never built for coffee-shop throughput. The Bitcoin blockchain can process around 7 transactions per second, compared to Visa Inc., which can process around 24,000 transactions per second. That ceiling is where Lightning picks up the load.
River Financial’s August 2023 analysis provides the clearest public evidence of actual usage growth: an estimated 6.6 million routed Lightning transactions in August 2023, equivalent to 2.5 transactions per second, compared to Bitcoin’s on-chain average of 4.4 TPS. Over two years, Lightning routed volume rose 546%, from K33’s August 2021 estimate of $12.1 million to approximately $78.2 million in August 2023.
The Capacity Paradox is the part most generic definitions miss. Public capacity has fallen from its 2024 highs while routed transaction counts keep climbing, because liquidity is moving into splicing transactions, private channels, and professionalized hub nodes that do not broadcast their state. Reading declining public BTC as a health signal is reading the wrong gauge.
1ML notes that Lightning Network statistics are approximations because nodes that do not broadcast their state are not included, which is exactly why River’s first-party operator data and on-chain capacity dashboards need to be read together, not in isolation.
Scale is one side of the ledger. Trade-offs sit on the other.
Pros, Cons, and Risks
Advantages
- Sub-second settlement: Settlement time for Lightning network transactions is under a minute and can occur in milliseconds, compared to Bitcoin’s ten-minute average block time.
- Tiny fees: A median base fee near 0.36 sats and a median fee rate of 0.000129 sat per sat transferred keep micropayments economically viable.
- Privacy-forward routing: Out of 5,923 public nodes, 3,010 run as Tor onion services, which shield node IP addresses from peers by default.
- High payment reliability in practice: River reported a 99.7% Lightning payment success rate in August 2023 across 308,000 transactions.
- Bitcoin-anchored finality: Every channel closes to the base chain, so the final balance inherits Bitcoin’s proof-of-work security.
Trade-offs and Risks
- Liquidity constraints: Each channel has a finite outbound balance, so large payments can fail to route even on an otherwise healthy network.
- Always-online requirement for penalty enforcement: Watching for a cheating counterparty broadcast requires either running a node or using a watchtower service.
- Capacity centralization pressure: With average channels per node at 6.35 and a 99th-percentile node capacity of 7.23 BTC, routing concentrates on larger hubs.
- Channel open/close costs: Funding and closing transactions still touch the base chain and pay on-chain fees, which can outweigh savings for rarely-used channels.
- Off-chain observability gaps: Private channels and direct payments are deliberately invisible, which makes accurate usage measurement genuinely hard.
Lightning’s trade-offs become clearer when held up against Bitcoin’s base layer.
Lightning Network vs On-Chain Bitcoin
The two layers are designed for different jobs. On-chain Bitcoin is an armored truck for final settlement. Lightning is the express lane for everything smaller and faster.
| Dimension | Lightning Network | Bitcoin On-Chain |
| Settlement time | Milliseconds to under a minute | ~10 minutes per block confirmation |
| Throughput (observed) | ~2.5 TPS routed (Aug 2023 estimate) | ~4.4 TPS average |
| Theoretical base-layer ceiling | Orders of magnitude higher (off-chain) | ~7 TPS |
| Median fee rate | 0.000129 sat/sat (≈0.013% of amount) | Variable, typically a few sats per vbyte |
| Finality model | Instant between channel peers; final on close | Probabilistic; stronger with each confirmation |
| Privacy default | Source and destination hidden from intermediate hops | Pseudonymous addresses, fully public transaction graph |
| Best use case | Micropayments, streaming, point of sale | Large-value settlement, long-term storage |
Source: 1ML, Lightning Network whitepaper, Wikipedia (for Bitcoin base-layer throughput context)
The key point for readers researching payment systems broadly: Lightning does not replace on-chain Bitcoin. It rides on top of it. Closing a channel is itself an on-chain transaction, so a working Lightning economy increases demand for base-layer block space rather than reducing it.
Real-World Applications
Retail Payments and POS
Coffee shops, convenience stores, and online merchants integrate Lightning to accept Bitcoin payments that settle faster than card networks and cheaper than credit-card interchange. River’s 2023 report identified gaming, social media tipping, and streaming as the primary use cases driving Lightning transaction growth, accounting for 27% of all growth.
Cross-Border Remittances
Lightning’s per-payment cost and speed make small cross-border transfers economic, where banks and money-transmitter fees would consume the principal. River’s 2023 report estimated monthly active Lightning users at between 279,000 and 1.116 million as of September 2023, with an estimated 1:8 ratio of non-custodial to custodial users, suggesting most everyday users reach Lightning via a hosted wallet app rather than by running a node themselves.
Streaming Sats and Gaming
Some apps charge by the second or by the message, creating payment flows that would be impossible on any chain where every transaction costs cents. The average Lightning transaction size on the public network was around 44,700 satoshis, or $11.84, in August 2023, and roughly a quarter of all payments fall in the 1-to-10-satoshi bracket tied to gaming and streaming.
Scenario: A Latte Paid in Sats
Imagine paying for a $4 coffee. Your wallet app quotes the invoice in sats. You scan the merchant’s QR code. Your Lightning client picks a route through two or three hub nodes already connected to both you and the merchant. Each hop locks the payment to a shared secret via an HTLC. Within a second, the merchant’s POS beeps, the invoice is marked paid, and the channel balances between the routing nodes have shifted by the corresponding amount. The Bitcoin blockchain never sees your transaction. It only sees the channel funding and, eventually, the channel closes.
The same flow repeats for streaming a video, tipping a writer, or paying a remittance, which is the practical bridge between this glossary entry and Bitcoin Lightning Network usage statistics that track how the mesh grows over time.
Frequently Asked Questions (FAQs)
Lightning is designed around cryptographic penalties rather than trust, but the safety profile depends on the entire stack, not just the protocol. Channels enforce cooperative behavior through revocable commitment transactions that let the honest party claim the full balance if a counterparty broadcasts an older state. Risks come from operational factors like being offline when a counterparty cheats, wallet bugs, and custodial trust if you use a hosted wallet.
You can still close the channel unilaterally by broadcasting the latest commitment transaction. After a timelock expires, your balance settles to your on-chain Bitcoin address. You do not need the counterparty’s cooperation to recover funds, but a cooperative close is cheaper and confirms faster than a forced close.
Public-channel data shows a median base fee near 0.36 sats and a median fee rate of 0.000129 sat per sat transferred, which works out to a fraction of a cent on a typical retail payment. Large payments that need to hop through liquidity-constrained channels can cost more, and opening or closing a channel still incurs a standard Bitcoin on-chain fee.
Lightning was designed specifically for Bitcoin and is deployed primarily on the Bitcoin mainnet. The three major implementations are Lightning Network Daemon (LND) by Lightning Labs, Core Lightning (CLN) by Blockstream, and Eclair by ACINQ, with a fourth library, Lightning Development Kit (LDK), maintained by Spiral, a subsidiary of Block, Inc. Some Litecoin nodes also run Lightning, and newer work on Taproot Assets extends Lightning to stablecoins issued on Bitcoin.
Conclusion
The Lightning Network settles Bitcoin payments in milliseconds at fractions of a cent by keeping routine activity off-chain and anchoring only funding and closing transactions to the base layer. As of early this year, 1ML shows 5,923 observable nodes, 18,807 channels, and 2,741.59 BTC of public capacity, while first-party operator data from River and others suggests real usage has outgrown what those public metrics alone reveal.
For a reader trying to judge the network’s health, the useful questions have shifted. Instead of asking whether public capacity is up or down this quarter, ask whether channel splicing is reducing the need for new public channels, whether routed volume and success rates are trending up, and whether the apps people actually use are adding Lightning rails. Those answers will matter more to the next phase of Bitcoin payments than any single snapshot of observable capacity.
