EigenLayer vs Ether.fi vs Kelp

Introduction

When you explore the restaking ecosystem in 2026, you will find three paths that were all sold to you as restaking: direct restaking on EigenLayer, Ether.fi's eETH and weETH, and multi-LST restaking through Kelp DAO (rsETH). One of them no longer is — Ether.fi is nearly finished unwinding its EigenLayer exposure. Each approach makes fundamentally different trade-offs between control, liquidity, risk, and yield, and choosing the right one depends on your specific priorities, capital size, and DeFi experience level.

You should understand that the differences between these protocols go deeper than surface-level feature comparisons. EigenLayer gives you direct control over operator selection and AVS opt-in decisions, but locks your capital behind withdrawal queues and offers you no DeFi composability. Ether.fi abstracts away the complexity of operator management and gives you a liquid token (weETH) that works across dozens of DeFi protocols, but you trust the protocol team to make sound operator and AVS allocation decisions on your behalf. Kelp DAO takes a different approach entirely, accepting multiple LST types as input and issuing rsETH, which provides you with diversified restaking exposure but adds another smart contract layer to your risk stack.

Each model serves a different user profile, and your optimal choice depends on whether you prioritise control, liquidity, simplicity, or diversification.

This comparison analyses all three protocols across the dimensions that matter most to you: TVL and market position, yield components and sustainability, risk profiles, DeFi integration depth, and fee structures. We provide you with concrete data and practical recommendations rather than generic overviews, helping you make an informed decision about where to allocate your restaking capital. The analysis reflects market conditions and protocol features as of August 2026, though the restaking landscape evolves rapidly and you should verify current figures before making your allocation decisions. For the foundational context on how restaking works, see our restaking explained guide.

You should recognise that understanding the differences between these three approaches is particularly important because your restaking decision is not easily reversible. Direct EigenLayer restaking involves a 14-day withdrawal escrow for you, and switching between LRT protocols requires you to sell one token and purchase another, potentially incurring slippage costs and tax events. Making an informed initial choice saves you both time and transaction costs compared to switching protocols after the fact.

Quick Comparison Table

Protocol Overviews

EigenLayer: The Restaking Infrastructure

EigenLayer is the base infrastructure layer that makes restaking possible. You should understand that it is not a liquid staking protocol — it is the smart contract system that allows staked ETH to secure additional services (AVS) beyond Ethereum consensus. Kelp DAO is built on top of EigenLayer and deposits its users' assets into EigenLayer's contracts; Ether.fi was, but has nearly finished withdrawing. If you restake directly on EigenLayer, you get full control over operator selection and AVS exposure but you receive no liquidity — your restaked ETH is locked behind a withdrawal escrow of 100,800 blocks, roughly 14 days, doubled from the original seven days when the slashing upgrade shipped.

EigenLayer supports two restaking paths for you: native restaking (requiring 32 ETH and a full validator with withdrawal credentials pointed to an EigenPod) and LST restaking (depositing liquid staking tokens like stETH or rETH into Strategy Manager contracts). If you choose native restaking, you get the purest form of restaking with no intermediary protocol fees, whilst LST restaking offers you flexibility with no minimum deposit. In both cases, you delegate to an operator who validates AVS on your behalf.

Your stake becomes slashable automatically once that operator opts into an AVS operator set — you do not consent per AVS, so monitoring your operator's allocations is your job. Slashing is live in code but barely exercised: on-chain, only about 20 AVSs have ever created a mainnet operator set, roughly 29 operators have registered into one, and 15 slashing events have ever fired. EigenLayer's burn address held no liquid staking tokens at all on 4 August 2026.

Ether.fi: Liquid Staking Behind a Restaking Label

Ether.fi is the largest protocol still marketed as liquid restaking — but the restaking part is being retired, and that is the most important thing to know before you deposit. Its own security documentation states that less than 1% of Ether.fi's assets remained restaked with EigenLayer as of August 2026, down from about half in early 2026, with the remainder due to go by Q3 2026 and EigenPod withdrawal credentials removed from its validators by Q4 2026. Ether.fi also states that its restaked ETH is not signed up to any slashable service and that, as policy, it does not sign up to slashable services.

What you are actually buying, then, is an Ethereum liquid staking token with a restaking label attached. Validators are run by node-operator clusters using distributed validator technology, with stakers retaining control over withdrawal credentials — that is the non-custodial claim, and it is about credential routing rather than any restaking feature. Ether.fi handles all staking complexity for you — validator setup, node-operator selection, reward accounting — behind a simple deposit interface, making staking accessible even if you lack technical expertise.

You receive two token variants from the protocol: eETH (a rebasing token whose balance increases as your rewards accrue) and weETH (a wrapped, non-rebasing version whose price appreciates relative to ETH). You should use weETH for most DeFi integrations because non-rebasing tokens are simpler to integrate into lending protocols, automated market makers, and yield tokenisation platforms. You can take confidence from the fact that Ether.fi has undergone multiple security audits from Omniscia, Certora, and Zellic, and maintains a bug bounty programme through Immunefi.

You should be aware that the core team still controls operator selection and protocol parameters. The role system that went live in July 2026 puts every contract upgrade behind a 6-of-10 multisig and a 10-day timelock with no emergency upgrade path, and bounds the share price with an on-chain 5% APR cap set by a 3-of-3 oracle committee.

Kelp DAO: Multi-LST Restaking

Kelp, now operating under the KernelDAO brand, takes a fundamentally different approach by accepting multiple liquid staking tokens (stETH, ETHx, sfrxETH) rather than staking ETH directly. When you deposit stETH into Kelp, it is restaked on EigenLayer and you receive rsETH in return. This multi-LST model provides you with built-in diversification across underlying staking protocols — your rsETH is backed by a basket of LSTs rather than a single protocol's validators. You should weigh the trade-off of less control over the underlying LST composition and lower TVL and liquidity compared to what Ether.fi offers you.

You benefit from the multi-LST architecture because your rsETH holdings are diversified across Lido (stETH), Stader (ETHx), and Frax (sfrxETH) validator sets. If one underlying protocol experiences a slashing event or smart contract issue, only the affected portion of your rsETH backing is impacted. However, you must also recognise that this architecture introduces additional smart contract dependencies — as an rsETH holder, you are exposed to vulnerabilities in Kelp's contracts, EigenLayer's contracts, and each underlying LST protocol's contracts simultaneously. Kelp is not custodial in the sense of a third party holding your assets: its NodeDelegator contracts create and own the EigenPods and hold the deposited LSTs in EigenLayer strategies, so validator withdrawal credentials point at protocol contracts rather than at an operator's wallet.

TVL and Market Position

When you evaluate TVL across restaking protocols, the direction of travel matters more than the level. DeFiLlama put EigenLayer at roughly $5.0 billion on 4 August 2026, against a peak above $22 billion in August 2025. That is a real outflow rather than an ETH price effect — restaked stETH alone fell from about $4.6 billion to roughly $0.5 billion over the same period. Treat the total as your broadest measure of how much economic security AVS networks can actually draw on, and note that it has shrunk by roughly three quarters in a year.

If you compare LRT protocols directly, Ether.fi leads at about $3.2 billion — roughly three to four times its nearest competitor — down from $12.4 billion in August 2025. Kelp DAO holds about $0.87 billion against a peak of $2.1 billion in September 2025. Ether.fi's scale still translates into deeper DEX liquidity for weETH and more integration venues, which is the practical reason to prefer it, but neither protocol is growing.

You should recognise that the TVL concentration in Ether.fi mirrors the broader pattern you see in liquid staking, where Lido dominates. This concentration creates a self-reinforcing advantage that benefits you as a depositor: more TVL means deeper liquidity, which attracts more DeFi integrations, which attracts more deposits. The practical implication for your portfolio is that weETH offers you significantly better liquidity and DeFi utility than rsETH or other competing LRTs. However, you must also consider that concentration means an Ether.fi-specific issue would affect a disproportionate share of your restaking exposure.

Market Concentration and Competitive Moats

When you assess market position stability, EigenLayer's role as the infrastructure layer is no longer unchallenged in the way it was in 2024. Competing restaking layers such as Symbiotic and Karak remain much smaller, but the more telling development is that the largest LRT has walked away from restaking entirely rather than switching to a rival.

If you choose Ether.fi, you benefit from strong dominance amongst LRT protocols, though it faces ongoing competition from Kelp, Puffer, and Renzo. If you choose Kelp, you should understand that its position is more vulnerable to competitive pressure, as its multi-LST model can be replicated and its TVL advantage over smaller competitors is narrower. You must consider protocol stability and competitive moat when making your long-term allocation decisions, as TVL shifts can directly affect the liquidity depth and DeFi integration availability you rely on.

Yield Comparison

Whichever protocol you choose, the same underlying Ethereum staking yield sits at the bottom of the stack, and it is thin: Lido displayed 2.2% APR on 4 August 2026, down from around 3% in early 2025 on DeFiLlama's stETH pool history. The restaking layer adds far less than the pitch implies. EigenLayer publishes no restaker APR, because rewards are denominated in EIGEN and AVS tokens rather than ETH. As an order-of-magnitude estimate only: about 50 million EIGEN a year goes to ETH and LST restakers under the year-one programmatic schedule, which at EIGEN's early-August 2026 price of about $0.185 is worth of the order of 0.2% a year against current restaked TVL, before the operator's cut. That schedule began accruing in 2024 and an Incentive Council created in March 2026 can change it.

EigenLayer direct: Your ETH yield is whatever your underlying staking route pays, around 2.2% APR in August 2026, plus those token rewards. You pay no EigenLayer protocol fee; operators set their own fee on AVS rewards anywhere from 0% to 100%, with 10% the default split. Two details are easy to miss: rewards stop accruing the moment you queue a withdrawal, and AVS reward distributions are published as a weekly Merkle root with a further week's delay before you can claim.

Ether.fi: weETH returned about 2.45% APR over the 30 days to 4 August 2026, measured from the on-chain weETH exchange rate — modestly ahead of wstETH's 2.20% over the same window. That figure is already net of the 10% Ether.fi takes from staking rewards (5% to node operators, 5% to the protocol). ETHFI incentives sit on top of it and are separate from the ETH-denominated return, so do not add them to a headline APR and treat the total as staking yield.

Kelp DAO: rsETH returned about 2.38% APR over the same 30 days, measured from the Kelp oracle's rsETH rate. The live incentive and governance token is Kernel under the KernelDAO brand, not the older KEP points token, and its emissions are smaller than Ether.fi's. The multi-LST model means your base yield is a weighted average of the underlying LSTs, slightly below a direct staking route because each underlying protocol's fee — Lido's 10% on stETH, for example — is deducted before Kelp receives anything.

In practice the realised yield differences between these routes were a few tenths of a percentage point in August 2026, well inside the noise of month-to-month variation. The restaking premium over plain liquid staking has effectively disappeared, which makes liquidity, DeFi composability, and risk profile the decisive factors rather than raw yield.

Yield Sustainability and Token Incentive Analysis

You must understand the critical distinction between organic yield (from base staking and AVS rewards) and token incentive yield (from ETHFI, Kernel, or other governance token distributions). Your organic yield is sustainable because it derives from protocol revenue — Ethereum consensus rewards and AVS service fees. Your token incentive yield is inherently temporary, funded by token inflation that dilutes existing holders and typically declines as protocols mature and reduce emission schedules.

If you hold weETH, Ether.fi's ETHFI programme has been the most generous amongst LRT protocols, but the sustainable, ETH-denominated part of your return is simply what Ethereum staking pays net of the 10% fee — about 2.45% APR in August 2026. Everything above that depends on a token price you do not control and on an emission schedule the issuer can change. If you entered Ether.fi primarily for token incentives, size the position on the 2.45% and treat the rest as optional.

If you choose Kelp DAO, the Kernel boost is smaller still, given the token's lower market capitalisation and trading volume. If you restake directly on EigenLayer, your return is EIGEN and AVS tokens on top of whatever your staking route pays — predictable in mechanism but not in value, since it is priced in tokens rather than ETH. None of the three routes currently offers a durable yield edge over holding a plain liquid staking token.

The wider lesson from 2026 is that the restaking premium was largely an incentive programme rather than a market for shared security. With realised net returns across weETH, rsETH and wstETH clustering within a few tenths of a percentage point, the non-yield differentiators — liquidity, DeFi composability, and the number of contract layers between you and your ETH — are what your decision should actually turn on.

Risk Profiles

EigenLayer direct — lowest smart contract risk, highest operational complexity: If you restake directly, you expose yourself to EigenLayer's contracts only (plus Ethereum staking contracts), avoiding the additional smart contract layer of LRT protocols. However, you bear full responsibility for operator selection, AVS evaluation, and position management. If you choose a poor operator, your returns and slashing exposure are directly affected. You need 32 ETH and validator infrastructure for native restaking; for LST restaking, you must understand the mechanics of delegation. Your operational burden includes monitoring operator performance, tracking AVS portfolio changes, manually claiming rewards, and managing the unbonding process when you want to withdraw. If you lack the time or expertise for active position management, you may find that this operational overhead outweighs the fee savings compared to LRT protocols.

Ether.fi — moderate smart contract risk, lowest operational complexity: When you use Ether.fi, you add its smart contracts (eETH/weETH tokens, liquidity pool, withdrawal queue, oracle) to the Ethereum staking contracts, increasing your total attack surface. By Ether.fi's own account there is no restaking slashing risk today, because none of its restaked ETH is enrolled with a slashable service; what remains is Ethereum validator slashing, smart contract risk, oracle risk and governance risk. You should note that operator selection is centralised — the Ether.fi team chooses operators on your behalf, which simplifies your experience but removes your individual control. Multiple audits from Omniscia, Certora, and Zellic should give you reasonable confidence in security.

You must also understand that centralised operator selection means all Ether.fi depositors share the same operator risk profile — if the team makes a poor selection decision, your weETH holdings are affected equally as those of every other depositor. This contrasts with direct EigenLayer restaking, where you can independently choose operators aligned with your personal risk tolerance and yield objectives.

Smart Contract Layer Comparison

Kelp DAO — moderate-high smart contract risk, moderate operational complexity: When you use Kelp, you add its contracts on top of EigenLayer, similar to Ether.fi. Additionally, you should understand that Kelp's multi-LST model exposes you to the smart contracts of the underlying LST protocols (Lido, Stader, Frax) in addition to Kelp and EigenLayer. This gives you the highest total smart contract layer count amongst the three options. The diversification benefit of holding multiple LSTs partially offsets your risk — if one underlying LST fails, only a portion of your rsETH backing is affected.

This is not theoretical. On 18 April 2026 an attacker forged a LayerZero cross-chain message that Kelp's rsETH bridge accepted, minting roughly 116,500 unbacked rsETH worth about $292 million — 2026's largest DeFi exploit, widely attributed to the Lazarus Group. The root cause was configuration rather than a contract bug: the bridge route had been set up with a single verifier, a 1-of-1 DVN. A DeFi United coalition led by Aave recapitalised the shortfall, with the final tranche of rsETH transferred on 25 May 2026 and backing restored above 100%.

Holders took no exchange-rate haircut. Kelp's oracle read about 1.0695 ETH per rsETH on 17 April, was frozen at 1.0696 from 19 April through the pause, and resumed at 1.0723 on 26 May — there is no downward step anywhere in the series, because the shortfall was absorbed by the recapitalisation rather than socialised into the rate. Redemptions were another matter: rsETH supply on Ethereum fell from about 634,600 in mid-April to about 435,800 on 4 August 2026. The exploit never touched EigenLayer's core contracts. It is bridge-configuration risk at the LRT layer, and that is what you are pricing when you size a Kelp position.

The custody question is worth settling, because Kelp is often described as custodial and its contracts say otherwise. Its NodeDelegator contracts create and own the EigenPods and hold deposited LSTs in EigenLayer strategies, so withdrawal credentials point at protocol contracts, not at an operator's wallet. Your real counterparty exposure is governance: Kelp's multisigs can pause the protocol and upgrade the contracts holding your assets. That is the same class of trust you extend to Ether.fi, and the one you avoid only in direct EigenLayer native restaking where you hold your own keys.

For a comprehensive analysis of these risks, see our liquid staking risks analysis.

Depeg Risk Comparison

Keep two ideas apart here, because conflating them is the most expensive mistake in this category. A liquid restaking token trading below its exchange rate is a market-price event; it is not a break in the redemption mechanism. Both weETH and rsETH are normally redeemable through their own protocols' withdrawal queues at the oracle rate, so a discount usually prices the time value of the queue and the depth of the DEX pool, not any impairment of the ETH behind the token. The exception is an incident pause. Kelp froze its core contracts on 18 April 2026 and rsETH mints, redemptions and rewards only resumed normal operation once the recapitalisation completed in late May — roughly five weeks with no redemption route at all. For Kelp the queue-time framing holds outside an incident, not during one.

If you hold weETH, you benefit from the deepest DEX liquidity amongst LRTs, which absorbs large sell orders with less price impact and keeps discounts shallow. The stress test worth studying is Ether.fi's own: over the 33 days from 18 April 2026, triggered by the Kelp incident, it processed 1,977 withdrawal requests totalling 542,792 ETH, or 19.6% of TVL. Median time to claimable was 4.9 days, the longest wait 16.7 days, and every request cleared within 17 days. The queue slowed; it did not fail.

If you hold rsETH, you face higher depeg risk due to thinner liquidity and the additional complexity of its multi-LST backing. You should be aware that if one of the underlying LSTs (stETH, ETHx, sfrxETH) experiences its own depeg event, your rsETH value is affected proportionally to that token's weight in the backing basket. This creates a cascading risk scenario in which your rsETH can depeg from both its own liquidity dynamics and underlying LST instability.

If you restake directly on EigenLayer, you have no depeg risk in the traditional sense — your position is denominated in ETH, not a tradeable token — but the 14-day escrow, plus the beacon-chain exit queue if you are natively restaking, means you cannot exit during market stress. That is a different form of liquidity risk, and it is the one you cannot hedge with a DEX sale.

DeFi Integration Comparison

EigenLayer direct: You get zero DeFi composability. Your restaked ETH is locked in EigenLayer's contracts and you cannot use it in any DeFi protocol. This is the fundamental limitation that LRT protocols were created to solve for you. If you need to use your restaked position as collateral, provide liquidity, or trade yield, direct restaking is not your option. You should also consider that the locked nature of your directly restaked ETH means you cannot respond quickly to market opportunities or risks — during a market crash, you cannot sell your restaked position or use it as collateral to hedge, whereas LRT holders can trade their tokens instantly on DEXs or deposit them into lending protocols for additional flexibility.

Ether.fi (weETH): You get the most extensive DeFi integration amongst LRTs. Your weETH is accepted as collateral on Aave V3, Spark Protocol, and Morpho Blue, though the loan-to-value parameters are set by each market's governance and change without notice, so check them before sizing a loop. You can use it in Pendle yield markets for fixed-rate and yield trading strategies. Curve and Balancer pools provide you with deep liquidity for weETH/WETH trading. This breadth of integration enables you to pursue sophisticated strategies including leveraged staking strategies, Pendle yield trading, and Curve LP provision.

You should value the depth of weETH liquidity on Curve because it means even your large positions can be entered and exited with modest slippage, which is critical during market stress when liquidity for smaller LRTs may evaporate. Pool depth is itself volatile — check the live pool before assuming an exit route exists at your size.

Integration Depth and Liquidity Analysis

Kelp DAO (rsETH): You get growing but more limited DeFi integration compared to weETH. Your rsETH is available on Pendle for yield trading and on select lending protocols, but with less depth and fewer venues than weETH. You should expect that your larger trades experience more slippage due to thinner DEX liquidity. The integration gap is narrowing as Kelp grows, but weETH's first-mover advantage and larger TVL provide a structural lead that you should not expect to close quickly. If you plan to employ leveraged restaking strategies or complex yield optimisation, you must factor in the liquidity differential when choosing between weETH and rsETH, as insufficient liquidity can prevent you from managing your positions efficiently during volatile market conditions.

Fees and Costs

EigenLayer direct: You pay no protocol fee. Your costs are only the operator's commission and Ethereum gas costs for your individual transactions. Operators set that commission themselves, anywhere from 0% to 100% of AVS rewards, and may vary it per AVS; 10% is the default split, so check your operator's actual rate rather than assuming a market norm. If you choose native restaking, you also bear the infrastructure costs of running your own validator. If you choose LST restaking, you pay the underlying LST protocol's fee (such as Lido's 10% on stETH), which is deducted before anything reaches EigenLayer, in addition to your operator commission on AVS rewards.

Ether.fi: You pay 10% of rewards in total, not 10% plus an operator cut on top: Ether.fi's documentation splits it 5% to node operators and 5% to the protocol, applied to both staking and restaking rewards. You also incur gas costs for depositing, wrapping (eETH to weETH), and any subsequent DeFi transactions. The 10% matches Lido's headline rate and is the price of the convenience and liquidity Ether.fi provides. It is already deducted from the roughly 2.45% APR quoted earlier.

Kelp DAO: Kelp's protocol fee is commonly reported at 10% of rewards, in line with Ether.fi's, but Kelp does not publish it prominently — treat that figure as unverified and check the live documentation before modelling returns. Additionally, if you deposit LSTs (stETH, sfrxETH), the underlying LST protocol's fee has already been deducted — meaning you effectively pay two layers of protocol fees (the LST fee plus Kelp's fee). This double-fee structure is the cost you accept for the multi-LST diversification model; with Kelp's own rate unpublished, though, the combined burden cannot honestly be stated as a number.

Fee Impact on Long-Term Returns

You should understand that the compounding effect of fee differences becomes significant over longer time horizons. At August 2026 yields the absolute amounts are small. Work from realised net rates rather than a synthetic gross-times-fee sum: over the 30 days to 4 August 2026 weETH returned about 2.45% APR, rsETH about 2.38%, and plain wstETH about 2.20%. On a $50,000 position that is roughly $35 a year between the two LRTs and about $125 a year between weETH and wstETH. Note also that a direct restaker pays no commission at all on the base staking yield — the operator's rate applies to AVS rewards only — so the headline fee percentages are not comparable cash flows.

However, you must weigh the fee comparison against the benefits each protocol provides you. Ether.fi's 10% staking fee buys you deep liquidity, extensive DeFi composability, managed validator operations, and automatic reward compounding. Kelp's double-fee structure buys you multi-LST diversification and simplified access to restaking if you already hold LSTs. Direct EigenLayer restaking gives you the lowest fees but requires the most technical expertise and provides you with no liquidity or DeFi utility. Your optimal fee-adjusted choice depends on how much value you derive from each protocol's non-yield features.

You should also factor gas costs into your total cost comparison, particularly if you hold smaller positions. Ether.fi and Kelp require you to make a single deposit transaction, whereas direct EigenLayer restaking requires separate approval, deposit, and delegation transactions. Reward claiming on EigenLayer is a manual process with per-claim gas costs you must pay, whereas LRT protocols handle your reward distribution automatically through token price appreciation. If your position is under $10,000, the gas cost savings from LRT protocols can offset a significant portion of their protocol fees for you.

Who Should Use What

Choose EigenLayer direct if: You have 32+ ETH and want to run your own validator with full granular control over operator and AVS selection. You should be technically proficient and comfortable managing validator infrastructure. You do not need DeFi composability for your restaked position. You want to minimise your smart contract exposure and protocol fees. You must be willing to accept the 14-day withdrawal escrow, and the beacon-chain exit queue on top of it if you are natively restaking.

Choose Ether.fi if: You want a liquid staking token with full DeFi composability and maximum liquidity for your position — but do not choose it for restaking exposure, because it is nearly finished unwinding that. You should choose this if you plan to use your staked position as collateral on Aave, trade yield on Pendle, or provide liquidity on Curve. You accept that validator credentials route to protocol contracts rather than an operator's wallet, whilst upgrades and pauses sit with Ether.fi's multisigs. You should prefer a simple deposit experience without managing operators or AVS yourself. You want the deepest LRT liquidity for easy entry and exit of your positions.

Choose Kelp DAO if: You already hold LSTs (stETH, sfrxETH, ETHx) and want restaking exposure without converting your tokens to ETH first. You should value the built-in diversification of the multi-LST model across multiple validator sets. You can accept moderate DeFi integration (less than weETH but growing). You want exposure to multiple underlying staking protocols through a single token in your portfolio.

Consider combining protocols: If you are a sophisticated investor, you should consider splitting your allocation across multiple approaches. You can follow a common strategy of 60% Ether.fi (for DeFi composability), 20% Kelp (for LST diversification), and 20% direct EigenLayer (for maximum control on a portion of your portfolio). This diversification reduces your concentration risk in any single protocol whilst maintaining your access to DeFi strategies through the Ether.fi allocation.

Position Sizing and Practical Considerations

You should understand that the minimum practical position size varies significantly across protocols. Direct EigenLayer native restaking requires holding 32 ETH (approximately $60,000 at ETH's early-August 2026 price of about $1,860) to run a full validator, making it accessible only to larger holders. LST restaking on EigenLayer has no protocol minimum for you but requires gas costs of approximately $15-25 for setup transactions, making positions under $5,000 inefficient from your fee perspective. Ether.fi and Kelp accept any deposit amount from you, with gas costs of approximately $10-20 for your deposit transaction.

If your position is under $10,000, you should clearly prefer LRT protocols — the gas cost efficiency of a single deposit transaction, combined with automatic reward compounding and DeFi composability, provides you with better capital efficiency than direct EigenLayer restaking. If your position is between $10,000 and $100,000, you can split between Ether.fi and either Kelp or direct LST restaking on EigenLayer for good diversification without excessive gas overhead. If your position exceeds $100,000, you should incorporate direct native restaking alongside your LRT allocations to achieve the broadest risk diversification and lowest fee burden on your directly restaked portion.

Conclusion

The restaking ecosystem offers you a genuine choice between control, convenience, and composability. If you restake directly on EigenLayer, you get maximum control and minimum fees but you sacrifice liquidity and DeFi utility. If you choose Ether.fi, you get the best combination of liquidity and DeFi integration, with validator credentials routed to protocol contracts, at the cost of centralised operator selection, multisig upgrade authority, and a 10% protocol fee. If you choose Kelp DAO, you get unique multi-LST diversification but with higher effective fees and less DeFi integration than Ether.fi provides you.

The honest framing for August 2026 is that this is no longer really a restaking decision. Ether.fi has almost finished exiting EigenLayer and holds no slashable service exposure, realised returns across all three routes sit within a few tenths of a percentage point of plain liquid staking, and EigenLayer's restaked TVL has fallen by roughly three quarters in a year. If you want the liquidity and DeFi utility, Ether.fi remains the most practical option — just buy it for what it is now rather than for a restaking premium that has largely evaporated. If you want genuine AVS exposure, direct EigenLayer restaking is the only route that still delivers it.

Regardless of which protocol you choose, the fundamental principles of restaking risk management apply equally:

• Slashing awareness: Separate Ethereum consensus slashing from AVS slashing, and check whether your operator has actually opted into a slashable operator set
• Operator monitoring: Track your operator's performance metrics, uptime history, and AVS selection decisions
• Smart contract exposure: Maintain awareness of smart contract risks across all protocol layers in your yield stack
• Position sizing: Size your restaking position appropriately relative to your total portfolio and risk tolerance

You should remember that the restaking ecosystem is still maturing, and the competitive landscape amongst LRT protocols will continue to evolve throughout 2026 — you should periodically reassess your allocation strategy as new data on protocol performance, security, and yield sustainability becomes available. To get started with liquid staking as a foundation for your restaking strategies, see our Lido referral guide and Rocket Pool referral guide.

Sources and References

• EigenLayer Official Documentation
• Ether.fi Documentation
• Kelp DAO Documentation
• DeFiLlama - Restaking TVL Data
• EigenLayer Review - CryptoInvesting
• Ether.fi Review - CryptoInvesting
• Liquid Staking Yield Strategies 2026 - CryptoInvesting

Frequently Asked Questions

Which restaking protocol has the highest TVL in 2026?

EigenLayer has the highest total TVL as the base restaking layer, at roughly $5.0 billion on 4 August 2026 per DeFiLlama, down from a peak above $22 billion in August 2025. Amongst liquid restaking token protocols, Ether.fi leads at about $3.2 billion, followed by Kelp DAO at about $0.87 billion. The old shorthand that LRT TVL is simply a subset of EigenLayer's no longer holds: Ether.fi's own documentation states that under 1% of its assets remained restaked with EigenLayer as of August 2026.

Is Ether.fi safer than direct EigenLayer restaking?

They now carry different kinds of risk. Ether.fi states that under 1% of its assets remained restaked with EigenLayer as of August 2026 and that its restaked ETH is not signed up to any slashable service, so its holders carry Ethereum validator, smart contract, oracle and governance risk rather than AVS slashing risk. Direct EigenLayer restaking keeps you on one fewer contract layer and gives you full control over operator selection, but it exposes you to AVS slashing and locks you behind a 14-day withdrawal escrow. Direct restaking is safer from a smart contract perspective; Ether.fi is safer from a slashing perspective today.

What is rsETH and how does Kelp DAO work?

rsETH is Kelp DAO's liquid restaking token. Unlike Ether.fi, which stakes ETH directly, Kelp accepts multiple liquid staking tokens (stETH, ETHx, sfrxETH) and restakes them on EigenLayer. This multi-LST approach provides built-in diversification across underlying staking protocols. When you deposit stETH into Kelp, it is restaked on EigenLayer and you receive rsETH representing your restaked position. rsETH can be used in DeFi similarly to weETH, though with less integration depth due to lower TVL and liquidity.

Which LRT has the best DeFi integration?

Ether.fi's weETH has the broadest DeFi integration amongst liquid restaking tokens. It is accepted as collateral on Aave V3, Spark Protocol, and Morpho Blue, integrated into Pendle yield markets, and available in Curve and Balancer liquidity pools. Kelp's rsETH has growing but more limited integration, primarily on Pendle and select lending protocols. Direct EigenLayer restaking has no DeFi composability — your restaked ETH is locked until you undelegate and wait through the 14-day withdrawal escrow.

Can I use multiple restaking protocols simultaneously?

Yes, you can diversify across restaking protocols by splitting your ETH allocation. For example, you might restake 50% of your holdings in Ether.fi (for liquidity and DeFi composability), 30% through Kelp DAO (for multi-LST diversification), and 20% directly on EigenLayer (for maximum control). Each protocol has different risk profiles and advantages, so diversification reduces concentration risk in any single protocol's smart contracts or operational decisions. Many sophisticated users employ this multi-protocol approach.

Financial Disclaimer

This content is not financial advice. All information provided is for educational purposes only. Cryptocurrency investments carry significant investment risk, and past performance does not guarantee future results. Always do your own research and consult a qualified financial advisor before making investment decisions.