Ethereum draft aims to cut finality from 16 to 4 minutes

Ethereum researchers have filed a draft proposal to make transactions final in about four minutes instead of roughly 16, CoinDesk reported. The plan was submitted on Wednesday by Ben Edgington, who coordinates Ethereum's fast-finality research, together with Francesco D'Amato, Roberto Saltini, Yann Vonlanthen and Mikhail Kalinin. It is unfinished and has no activation date.

What "finality" means

A transaction can appear in an Ethereum block within seconds, but it only becomes final once validators — the computers that secure the network under proof of stake — have committed never to rewrite it. Today that takes about 16 minutes. Anyone releasing money on another chain against an Ethereum deposit has to decide whether to wait for that point or take the risk.

Why 16 minutes costs money

Fast cross-chain services usually pay the customer immediately from funds they already hold on the destination chain, then wait for Ethereum to settle before they are repaid. That float is idle capital. In CoinDesk's example, a service fronting a steady $1 million an hour has about $267,000 waiting on repayment with a 16-minute delay, but only about $67,000 at four minutes.

Stablecoin issuers already price this in. Circle charges 0.01% ($1 per $10,000) for its fast USDC route out of Ethereum, before network and app fees, while its standard route has no Circle transfer fee. Circle also caps the value of fast transfers awaiting settlement, so quicker finality frees capacity sooner. Our explainer on how crypto bridges work covers why these services must wait at all.

How the proposal would work

• Today each validator casts one vote that does two jobs: helping pick the next block and helping lock older blocks in for good.
• Validators vote in rotation, and a block is final only after two full rounds of everyone voting, so finality can move no faster than that cycle.
• The draft gives each job its own vote, letting finalisation run on a separate, faster clock that developers could keep tightening.
• The researchers say the split would also make it easier to adopt quantum-resistant signatures later — the same long-term concern behind Starknet's post-quantum research vote.

What is still open

Voting timings, penalties and parts of the security analysis remain unresolved; Edgington wrote that "many aspects remain TBD". The immediate goal is to obtain a formal proposal number, which puts the idea into Ethereum's upgrade pipeline. Even four minutes would lag Solana, which switched on Alpenglow on Wednesday with a target of about 150 milliseconds.