Bitcoin's base layer processes only a handful of transactions per second, and fees rise when blocks are full. The Lightning Network, proposed by Joseph Poon and Thaddeus Dryja in a paper first drafted in 2015 (final version January 2016), is a second layer built to fix that.
Payment channels
Two users lock bitcoin into a shared 2-of-2 multisig output on-chain. This opens a payment channel. They can then update the balance between them as often as they like, instantly, by signing new transactions they keep off-chain. Only the opening and closing hit the blockchain.
A network of channels
You don't need a channel with everyone. Payments are routed across connected channels, hop by hop. Hashed time-locked contracts make sure a payment either completes fully or not at all — no middleman can steal it.
Benefits
- Near-instant settlement
- Fees often a fraction of a cent
- Payments as small as one satoshi (even fractions, tracked in millisatoshis)
Trade-offs
Channels need liquidity, your node should be online (or use a watchtower) to stop cheating, and routing large payments can fail. Many users choose simpler custodial Lightning wallets, giving up some self-custody.
Lightning is now used for tipping, remittances and point-of-sale payments, showing how Bitcoin can scale in layers.