Finality is the moment a blockchain treats a transaction as permanent. After this point, the record cannot be changed, reversed, or canceled, which keeps the ledger trustworthy and tamper-resistant.
Finality prevents double-spending and gives users a clear settlement. It also underpins the reliability of dApps that need predictable outcomes once a block is finalized. These guarantees build user trust in a system with no central operator.
Traditional payment systems often rely on banks and clearinghouses that can reverse or amend entries. Blockchains aim for the opposite outcome, using decentralized consensus so that once a transaction is finalized, it stays put. This contrast explains why finality matters so much for crypto payments and smart contracts.
Transaction finality is about one specific transfer becoming irreversible. State finality goes wider and locks in the whole chain’s current state, which is handy for apps that depend on many contracts staying in sync.
Security grows as more blocks are built on top of a transaction. Bitcoin is the classic case, where many users treat six confirmations as “final,” which roughly aligns with about an hour of block production.
Some networks reach a firm decision at the moment validators agree, leaving no waiting period for extra confirmations. Systems using Byzantine fault-tolerant algorithms, like Tendermint or Ripple’s consensus, aim for this style of finality.
Proof-of-stake chains can “seal” checkpoints when a supermajority of validators attests. Reversing those blocks would require burning a large staked amount, which makes rollbacks uneconomic. Ethereum’s post-Merge design is a common example.
Some definitions highlight a stricter notion where, once finalized, a transaction cannot be rolled back under any circumstances.
In a few protocols, finality refers to the chain’s global state becoming irreversible after a state transition, not just a single transaction or block.
Time-to-finality depends on the consensus design and on network latency. Low-latency networks can finalize faster, while others trade speed for other goals. On Bitcoin, the common six-confirmation convention maps to about 60 minutes.
Forks and temporary chain reorganizations delay confidence in probabilistic systems. Latency can slow down agreement even in fast designs. Proof-of-work networks also face 51 percent attack risks, where enough mining power could try to rewrite short-range history. These issues explain why many newer systems focus on stronger or faster forms of finality.