Staking Router v3 (SRv3) is a Lido Core upgrade that supports large validators (up to 2048 ETH) and introduces balance-based accounting for the post-Pectra validator model, lays the foundation for proposed new versions of the Curated and Community Staking modules, and enables efficient cross-module stake migration. More details can be found in LIP-35.
This proposal seeks Lido DAO approval for the SRv3 architecture, including the consolidation-based migration approach, key parameters, and related operational controls.
Before Pectra, every Ethereum validator had a fixed effective balance of 32 ETH. EIP-7251 has invalidated the fixed-balance assumption: a validator’s effective balance can now range from 32 ETH to 2048 ETH, with rewards scaling according to validator size. As a result, a validator at the new maximum can earn up to 64 times as many rewards as a 32 ETH validator. The current Lido accounting treats all validators identically — as equivalent units — regardless of their effective balance. This validator-count-based model remains functional as long as the Lido protocol only supports 0x01 withdrawal credentials and all Lido validators remain effectively capped at 32 ETH. However, for large validators with max effective balance 2048 ETH, it would misallocate stake between modules by relying on validator count rather than actual balances, and fail to accurately track the total amount of ETH backing stETH.
At the same time, the protocol needs a reliable way to migrate stake from Curated Module v1 (CMv1) to the new Curated Module v2 (CMv2) introduced in LIP-33 (subject to DAO approval), and to make ETH available for the migration regardless of withdrawal demand. The current stake rebalancing mechanism (via organic deposits and withdrawals) is too slow to support a coordinated migration.
SRv3 addresses these issues and lays the foundation for native support of large validators and future consolidation-based stake migration between other modules.
Staking Router v3 is the foundational infrastructure upgrade that serves as the base layer upon which LIP-33 (CSM v3 and CM v2) is built. The proposed SRv3 upgrade reworks five areas of Lido Core:
0x02 validator top-ups up to 2048 ETH;The full specification, including contract interfaces, migration procedure, and edge cases, is available in LIP-35. The full set of proposed module parameters is published on IPFS.
The SRv3 codebase is currently undergoing security audits of on-chain code by Certora and Statemind, voting script and the Easy Track contracts by MixBytes, and off-chain Oracle code by Composable Security. Final audit reports will be published on the Research forum prior to the mainnet on-chain vote for the SRv3 release.
SRv3 introduces a set of operational controls to ensure safe and controlled functioning of the upgraded staking system:
If this proposal is approved:
SRv3 establishes the core infrastructure required for LIP-33 (CSMv3 and CMv2), including balance-based accounting and consolidation-based workflows. As such, both proposals are intended to be adopted and implemented in a coordinated manner. If the proposals are not both approved in the same voting cycle, the rollout process for both upgrades will need to be reassessed.