Rocket Pool (RPL, rETH): permissionless Ethereum staking and Saturn 2

Holders of rETH are getting a protocol-level way to cash out. On 13 October 2026 the node operators who govern Rocket Pool, Ethereum’s fourth-largest liquid staking protocol, closed four signalling votes on RocketDash: the protocol DAO put rETH withdrawals into the planned Saturn 2 upgrade, picked random selection when megapool validators must be exited, kept the operator bond at 4 ETH and left on-chain signal voting out. Saturn 2 itself still needs a final package vote, and its mainnet launch is only a target, May 2027. Until then, rETH holders who want ETH back still go through the market.

Rocket Pool Saturn 2 votes: what passed on 13 October 2026

Voting ran on RocketDash, the site the Rocket Pool docs name for off-chain pDAO votes, from about 05:50 UTC on 29 September to about 06:00 UTC on 13 October 2026. Quorum was 9,883 units of vote power out of 65,886. All four questions cleared it. Turnout, counted in addresses, was small: 40 to 42 per vote.

Source: RocketDash governance API, read on 14 October 2026 at about 18:15 UTC.

rETH withdrawals are in

RPIP-71 passed with 97.6% of the vote power cast. The proposal, written by the community contributor knoshua and created in May 2025, starts from a plain admission in its abstract: rETH “is fully backed by staked ether but currently lacks a direct mechanism for stakers to signal a desire to exit.”

The mechanism is a queue. A holder parks rETH in it, and from that moment the parked rETH stops earning. The queue is served before any new validator gets ETH. Once requests pass a threshold, the protocol forces validators to exit using EIP-7002, the Ethereum change that lets a contract on the execution layer trigger a validator exit. Redemptions therefore wait on Ethereum’s own exit queue. They will not be instant.

Who gets exited is the touchy part. RPIP-71 says the Oracle DAO should pick older minipools first, preferring those with the highest commission, and that megapool validators should be chosen by a rule nobody can game. This vote set that rule to random. The rival option, starting with validators furthest below their bond requirement, lost 8,507.2 to 5,855.4 in the last round.

The minipool rule is not settled. Governance analysts DAOplomats posted simulations on the forum on 7 October 2026, from a snapshot taken on 5 October: 11,571 minipools across 1,323 nodes were eligible, and the 14% commission tier alone held 58.11% of them. Their post notes that minipool exit ordering stays open to a later RPIP. Feelings run high among operators who would be exited: one long-time minipool operator wrote in the thread on 2 September that they would vote against RPIP-71 and leave the protocol if it passed.

The 4 ETH bond stays

RPIP-86, the living overview of Saturn 2, framed the bond question as three choices: lower it to 1.5 ETH, keep 4 ETH or raise it to 6 ETH. The 1.5 ETH path came from RPIP-42, passed in 2024, which planned 1.5 ETH bonds for each validator after a node’s first two. It drew 170.8 units of vote power. RPIP-83, which would have set 6 ETH per megapool validator and used rewards to top bonds back up, drew 1,907.5. Holding at 4 ETH won with 12,174.9.

Lower bonds would only add operators, and Rocket Pool already has more of them waiting than rETH deposits to pair them with. knoshua summed it up on the forum on 3 August 2026: “The rETH demand simply isn’t there yet.” The same post noted that the bond can be lowered later with a settings change if demand shows up.

On-chain signal voting is out

RPIP-85 wanted to move signalling votes on-chain. Cost was part of the case: the proposal says Snapshot Pro, which the old voting setup relied on, costs $600 a month. It got no votes in favour; 11,883.0 units voted against and 2,370.0 abstained.

What still needs a vote

Nothing is final yet. On 29 September 2026 knoshua set out four steps on the forum: sentiment polls, the four RocketDash votes that have now closed, a sentiment poll on the whole package, then a RocketDash vote on Saturn 2 as one package. The last two have not happened. No date for the package vote had been published by 14 October 2026.

Saturn 2 also carries changes that were not on this ballot. According to RPIP-86, anyone could challenge validators with poor attestation records and have them ejected (RPIP-73). It adds an exit-request framework with minipool penalties (RPIP-80), faster withdrawal-credential checks (RPIP-79), pDAO spending of ETH and ERC-20 tokens (RPIP-82) and revised governance guardrails (RPIP-87). It would also cut RPL issuance to 2.5% a year and end RPL rewards for node operators. Two items from the 2024 plan were pushed back: 1.5 ETH bonds, and a richer revenue share for RPL such as burning it or buying it to add liquidity.

What is Rocket Pool?

Rocket Pool is a liquid staking protocol for Ethereum with two kinds of users. Stakers deposit ETH and get rETH, a token that represents the staked ETH plus rewards. Node operators, who want to run a validator but do not have 32 ETH, post part of it themselves and borrow the rest from those deposits.

Anyone can be an operator. There is no application form and no whitelist: you post a bond and run the software. Lido, the market leader, picks its operators instead. In its bi-weekly update for 6 October 2026, published on 12 October, the Rocket Pool team counted 1,414 node operators and 18,143 pending or active validators.

New to how staking works? Ethereum pays validators for proposing and attesting to blocks and penalises them for going offline or breaking the rules. That is the proof-of-stake model Ethereum switched to at the Merge on 15 September 2022.

How Rocket Pool staking works

rETH for stakers

The minimum deposit is 0.01 ETH, according to the Rocket Pool docs. Deposits go through the official staking app; rETH can also be bought on decentralised exchanges. The token does not rebase. A wallet holds the same number of rETH over time, and each one is worth more ETH as rewards come in. On 14 October 2026 at about 18:20 UTC the rETH contract reported 1.17379 ETH per rETH, with about 314,953 rETH in circulation.

The docs say rETH can be swapped back to the protocol for ETH “if there is liquidity available”. On 14 October 2026 there was almost none. The rETH contract held a few billionths of a billionth of an ETH, read on-chain, and the deposit pool held 31.74 ETH, according to RocketDash. Against roughly 370,000 ETH worth of rETH, that is nothing. Leaving rETH today means selling it on a decentralised exchange or another market, at whatever price that market offers.

DefiLlama put the rETH yield at 2.15% a year on 14 October 2026, with a 30-day average of 2.16%. Stakers get what is left of validator rewards after the operator’s 5% base commission and the 9% voter share paid to RPL stakers, according to the Saturn 1 docs.

Rocket Pool node staking: megapools and the 4 ETH bond

A node operator bonds some ETH and borrows the rest of a 32 ETH validator from the deposit pool. It earns rewards on its own ETH plus a commission on the borrowed ETH. The bond is what protects stakers: a validator that is penalised or slashed eats into its operator’s ETH, and the Rocket Pool FAQ says penalties and losses are socialised across the network rather than landing on one user.

The bond has shrunk with each upgrade. It was 16 ETH at the November 2021 launch. Atlas cut it to 8 ETH on 18 April 2023. Saturn 1, live since 18 February 2026, brought megapools and a 4 ETH bond, with 28 ETH borrowed from stakers. A megapool is a single contract per operator that can hold many validators. The Saturn 1 docs link the two changes: “Megapools reduce operational cost, allowing Rocket Pool to lower the bond requirement per validator from 8 ETH to 4 ETH.”

For operators the arithmetic is attractive. At a 5% commission, the docs calculate, an 8 ETH minipool earns the equivalent of 15% commission per ETH bonded, while a 4 ETH megapool validator earns 35%. Since Saturn 0 in October 2024, operators no longer have to hold RPL to join.

That is why they queue. New megapool validators wait until there is staker ETH to match them, and on 14 October 2026 RocketDash counted 2,345 of them waiting (1,558 in the express queue, 787 in the standard one).

The Oracle DAO and the protocol DAO

Two governance bodies run the protocol. The Oracle DAO (oDAO) is an invite-only group of nodes that carries Beacon Chain data, such as validator statuses and balances, over to Ethereum’s execution layer. It also checks new validators’ withdrawal credentials, builds the rewards Merkle tree and polices fee recipients. Coinbase said in August 2023 that it takes part in the oDAO. Stakers have to trust this group to report honestly, an assumption that sits awkwardly inside an otherwise permissionless system.

The protocol DAO (pDAO) is made up of node operators, who vote with power based on staked RPL. Houston, live on 17 June 2024, brought on-chain proposals and a Security Council for urgent action; the protocol’s own timeline dates the launch of on-chain pDAO governance on mainnet to 20 November 2024. Saturn 1 added guardrails under RPIP-60: a delay between an upgrade passing and being executed, and a Security Council veto.

The Smoothing Pool

Execution-layer income such as priority fees and MEV is lumpy. One validator lands a valuable block; another waits weeks for any block at all. Operators who opt into the Smoothing Pool, added with the Redstone upgrade in August 2022, pool that income in a shared contract that splits it evenly at the end of each 28-day rewards interval. Those who stay out get a Fee Distributor contract of their own. RocketDash counted 3,081 nodes in the Smoothing Pool on 14 October 2026.

Rocket Pool in numbers, October 2026

The direction is down. On DefiLlama data retrieved on 14 October 2026, TVL fell from $2.605 billion on 15 October 2025 to $1.273 billion, after peaking at $3.172 billion on 14 August 2025. Counted in ETH the fall is smaller, from 634,688 ETH to 521,952 ETH, or about 18%; the rest is the lower ETH price. rETH supply slipped 0.2% to 316,419 in the update dated 22 September and another 0.3% to 315,605 by 6 October, according to the team. Operators kept queuing all the while.

RPIP-86 states the aim of RPIP-71 in those terms: a liquidity guarantee should make rETH more attractive and keep it from trading below its peg for long stretches. More rETH demand is what would empty the operator queue.

Rocket Pool vs Lido: rETH and stETH

By DefiLlama’s count on 14 October 2026, Ethereum liquid staking protocols held about $41.46 billion. Lido had $23.65 billion of it (57.0%), Binance staked ETH $8.59 billion (20.7%), ether.fi $4.66 billion (11.2%) and Rocket Pool $1.27 billion (3.1%), in fourth place.

The tokens work differently. stETH rebases: Lido’s documentation says balances are normally recalculated daily when its oracle reports. rETH keeps a fixed balance and rises in value against ETH, as Lido’s wrapped wstETH does. Who runs the validators differs too: a curated list for Lido, anyone with a 4 ETH bond for Rocket Pool. Lido’s size buys deeper markets for its token, and that gap matters while rETH has no protocol exit.

The RPL token: from collateral to revenue share

RPL is an ERC-20 token on Ethereum, at contract 0xD33526068D116cE69F19A9ee46F0bd304F21A51f. It launched in 2017 with a fixed supply of 18,000,000. In 2021 holders migrated to a new contract with 5% yearly inflation, and on 14 October 2026 the on-chain supply was 22,875,010.84 RPL. The old token (0xB4EFd85c19999D84251304bDA99E90B92300Bd93) still exists. Do not confuse the two.

Until Saturn 0 in October 2024, operators had to stake RPL worth at least 10% of the ETH they borrowed. Saturn 0 made it optional. Since Saturn 1, RPL staked on megapools earns a 9% “voter share” of protocol ETH revenue under the Universal Adjustable Revenue Split, and it still carries voting power. Operators take a 5% base commission, the pDAO share is set at 0%, and the rest goes to rETH holders.

Inflation is now split 50% to node operators, 47.5% to the pDAO and 2.5% to the oDAO. That split comes from RPIP-81, which passed on Snapshot on 19 June 2026 with 11,735 votes for and 100 against. Its motivation section begins: “Rocket Pool currently faces a protocol funding shortfall.” The Grants Management Committee has been swapping RPL for USDC over the counter every month; the October 2026 swap was for $1,500. Under Saturn 2 as planned, issuance falls to 2.5% a year, with 95% to the pDAO, 5% to the oDAO and nothing to node operators.

This page describes how the token works. It is not a recommendation to buy or sell RPL or rETH.

Rocket Pool history: from a 2016 idea to Saturn 1

• December 2016: David Rugendyke starts work on Rocket Pool.
• 19 May 2017: alpha release on GitHub.
• September 2017: RPL presale, 8,220,000 tokens. A public crowdsale of 7,080,000 RPL followed later in 2017. The ETH raised was not published.
• May 2018: Consensys Ventures invests, according to Rocket Pool’s own timeline. The amount was not disclosed.
• July 2018: the first public beta starts. Five public betas followed, with more than 100,000 testnet ETH staked.
• October 2021: StakeWise co-founder Dmitri Tsumak discloses a withdrawal-credentials bug affecting Rocket Pool and Lido. Launch is postponed while the fix is written and reviewed by Sigma Prime.
• 9 November 2021: mainnet launch with 16 ETH minipools.
• 29 August 2022: Redstone, with the Smoothing Pool, ahead of the Merge.
• 18 April 2023: Atlas brings 8 ETH minipools and support for Shapella withdrawals.
• 10 August 2023: Coinbase Ventures buys an undisclosed amount of RPL; Coinbase says it runs nodes and takes part in the oDAO.
• 17 June 2024: Houston launches, with on-chain pDAO proposals and a Security Council.
• 28 October 2024: Saturn 0 removes the RPL requirement.
• March 2025: RPIP-65 passes, sending new deposits to an rETH withdrawal buffer sized at 1% of rETH TVL.
• 18 February 2026: Saturn 1 goes live with megapools and 4 ETH bonds.
• 13 October 2026: the Saturn 2 signalling votes close.

The 2021 bug is the closest the protocol has come to a security incident. It was fixed before launch, and checking withdrawal credentials is now an oDAO duty. DefiLlama lists no hack of the protocol.

Saturn 2 timeline: devnet in October 2026, mainnet target May 2027

The core team posted an indicative plan on the DAO forum on 10 August 2026, with three development networks before audits:

• Devnet-1, October 2026: ejections, exit requests, faster credential checks, execution-layer triggered exits.
• Devnet-2, January 2027: rETH withdrawals (RPIP-71) and pDAO spending.
• Devnet-3, February 2027: bond curves, the revenue split and the remaining items.
• Audits and testnet: target April 2027.
• Mainnet launch: target May 2027.

The same post carried a caveat: “Audit timelines have historically been difficult to predict, and we still need to engage auditors for this phase.” Two Devnet-3 items, on-chain voting and RPIP-83, have since been voted out. The team’s update for 6 October said the first devnet deployment was expected towards the end of the month.

Rocket Pool security: audits, track record and risks

In almost five years on mainnet the protocol’s contracts have not been exploited. Sigma Prime, Consensys Diligence and Trail of Bits reviewed the code in 2021, before launch. Each later upgrade went through audits of its own: Redstone in June 2022, Atlas between December 2022 and January 2023, Houston between December 2023 and April 2024, and Saturn 1 in January and February 2026 with Cantina, Bailsec, Sigma Prime and Hashlock. An Immunefi bug bounty pays up to $150,000 for critical bugs.

The open risks are elsewhere. rETH has no working in-protocol exit today, so a holder who needs ETH depends on market prices. The Oracle DAO, a small invited group, reports the numbers everything else relies on. Saturn 2 is not audited, auditors had not been engaged as of August, and its schedule is a target. The pDAO says it is short of funds, and RPL is a small, inflationary token. Turnout is thin: about 40 addresses decided this month’s votes. And smart contract risk never drops to zero; our guide to DeFi and liquid staking risks covers it in more detail.

Next on the calendar: the first Saturn 2 devnet, expected in late October 2026, then the sentiment poll and RocketDash vote on the full package, neither of which had a date on 14 October.