Ethereum Draft Would Let Validator Deposits Carry 8,192-Byte Keys

A draft proposal filed to Ethereum's improvement-proposal repository would rebuild the contract that admits new validators, so that it can accept signature keys far larger than the ones the network uses today. The pull request, number 12235, is titled "Post-quantum-ready deposit contract". It swaps the hard-coded fields of the current design for variable-length ones. Nothing in it picks a post-quantum algorithm. That choice is pushed to a separate change at the consensus layer.

What the contract stops hard-coding

The deposit contract is the entry point for staking: a would-be validator sends ether to it with a public key, and the consensus layer picks the deposit up from there. The current version bakes in BLS12-381, the signature scheme Ethereum has used since the Beacon Chain went live. Public keys sit at 48 bytes, signature and credential metadata at 96. COINOTAG carries that pair. The draft's ceiling, reported by both COINOTAG and Decrypt, is 8,192 bytes for each field.

The extra room is not arbitrary. Decrypt notes that stateless hash-based signatures standardised by NIST run at roughly 8KB, though it is the only account to draw that link. Each deposit would also carry a scheme identifier showing which cryptography it uses. Scheme 0 is reserved for BLS. Later numbers are left open for systems Ethereum has not chosen.

A one-way switch and the half it leaves out

Three states are defined for the contract: disabled, BLS enabled, and BLS retired. Retirement is irreversible. Once it is activated, new BLS deposits stop for good, while validators already running on BLS keys keep operating and are not forced to move. Migration would route through EIP-7685, the general request system between Ethereum's execution and consensus layers, alongside the existing rules for deposits, withdrawals and validator consolidations.

What the proposal does not do is written into it.

The draft excludes "existing keys, exits, credential changes, and migration into a post-quantum scheme", a set it calls "the harder half of the problem".

Two counts of the stake behind BLS

Here the reporting splits. CoinDesk's Wednesday piece and COINOTAG both put the staked ether secured by BLS keys at 42.4 million, worth about $104 billion at that day's prices. Cryptobriefing, publishing the same day, wrote roughly 37 million ETH across around a million validators. The gap between them is 5.4 million ETH. Neither names where its figure came from. We could not establish which one supersedes the other, so both stand here unreconciled.

Nothing has been scheduled

The proposal has no assigned number. Several outlets report the suggested designation as EIP-8394, while the repository bot's template comment in the pull request still points at the placeholder eip-9999, and Decrypt calls the draft "EIP-12235", which is the pull-request number. The filing date is unsettled too. The Crypto Times dates publication to 24 August, CoinDesk covered it on 26 August, and Decrypt's account places the submission on "Monday, August 26", a day that fell on a Wednesday.

The labels on the pull request read draft, waiting on editor review, waiting on editor consensus, waiting on CI. CoinDesk reports the Ethereum Foundation aiming at roughly 2029 for the core protocol work; The Crypto Times writes that full post-quantum consensus is targeted toward 2030. No fork, testnet activation or mainnet deadline has been named for this draft, and its contract address and activation timestamps are undecided. Banks have already run post-quantum transfers on a separate testnet this month. Ethereum's version of the question starts as a working paper with failing automated checks.