Jito vs Marinade: Solana LST Fees, MEV and APY Compared

By DifiCalc Research Team · Published Sep 18, 2026 · Reviewed Sep 18, 2026

TVL, APY and fee figures reviewed Sep 18, 2026 against official documentation and dashboards; yields move each epoch — verify current numbers before staking. See our review methodology.

Two Solana staking pioneers

Marinade was first. When it launched in 2021, Solana staking meant locking SOL with a single validator through an epoch-long cooldown. Marinade pooled delegation and issued mSOL, a token whose value in SOL rises continuously as inflation rewards compound — the first liquid staking primitive on the network, and still the most widely integrated across Solana lending and DEXs. It also offers Marinade Native, a no-token product that delegates directly through its automated strategy, for holders who do not want an LST at all. More background sits in our Solana staking vs lending guide.

Jito arrived from the MEV side. Its validator client runs an auction in which traders bid for transaction ordering, and JitoSOL, launched in late 2022, routes stakers into validators capturing those bids. Its rewards therefore have two engines — ordinary inflation and MEV tips — and the JTO governance token backs the ecosystem around the client, the TipRouter and the newer Block Assembly Marketplace. The trade-off is structural: JitoSOL holders are aligned with Jito's software roadmap. Our full assessment is in the Jito protocol review.

Fees: reward-based vs exit-based

The fee models now point in opposite directions. JitoSOL keeps a traditional asset-management fee: 4% of all rewards earned, applied after validator commissions, which works out to roughly 0.3% of deposited value per year; selling JitoSOL on a DEX like Jupiter avoids the separate 0.1% direct-unstake charge and the up-to-one-epoch wait. Marinade took the other path: its long-standing 6% performance fee was removed, and current core SOL products keep 100% of epoch rewards, with revenue collected on the way out. That makes headline APY cleaner but shifts cost to anyone leaving.

Fee schedule side by side

MEV: auctions, BAM and Marinade's SAM

Jito's edge is its ordering auction. Searchers bid tips to validators running the Jito client; winning bids are distributed transparently through the TipRouter — which retains 3% — and JitoSOL's share accrues into its price. Validators set their own MEV commission, typically inside a 0–10% band, and StakeNet's ranking explicitly penalizes high-commission operators over a 30-epoch window. The roadmap now pushes toward BAM, a block assembly marketplace, with stake increasingly directed to BAM validators and the DAO capturing 5.7% of tips plus growing coordination revenue. The trade-off: as Jito client adoption exceeds roughly 60% of network stake, client and infrastructure concentration becomes a network-level debate.

Marinade's v2 answers with different machinery. The Stake Auction Marketplace lets validators bid for delegated stake, pushing commissions — including MEV terms — down competitively rather than routing through one client, while Protected Staking Rewards posts validator bonds that pay stakers for missed rewards during downtime. It does not capture the same magnitude of MEV tips in normal conditions, but it keeps delegation open to validators regardless of client. Our overview of MEV and user protection explains why transparent distribution matters.

APY behavior: lumpy vs steady

JitoSOL's total return is a stack: staking inflation of roughly 5–7% plus MEV contributions that commonly add 1–2% but vary enormously — high-volatility days with liquidations and arbitrage pay well; quiet, range-bound weeks pay almost nothing. That makes any single-epoch APY misleading and explains why Jito can lead for a quarter and then converge. Marinade's return is smoother: broad delegation smooths out individual validator performance, and without a performance fee the full inflation reward lands in mSOL's exchange rate, though its MEV capture is smaller.

Both tokens also support yield stacking: mSOL and JitoSOL can be deposited into lending markets or liquidity pools, where the LST's own APY compounds under protocol fees — a strategy with its own depeg and liquidation risks, covered in our LST yield stacking guide. If you want the category leaders ranked, see our best liquid staking ranking.

The wider Solana LST field: Sanctum, Coinbase and switching costs

Jito and Marinade no longer operate in isolation. Sanctum, the third notable Solana staking system, runs an LST router that lets wallets and aggregators swap between many different LSTs through unified liquidity, alongside its own LSTs and a strategy that spreads stake across the network. For users it lowers switching costs — moving between JitoSOL and mSOL becomes a routed swap rather than a full unstake — and for the protocols it creates a shared liquidity layer that keeps both tokens deeper than their direct withdrawal mechanisms would allow. The trade-off is another dependency: router outages or imbalanced LST pools can briefly distort prices, and an LST routed through several wrappers is harder to analyze than the original token.

Coinbase and other custodial venues sit at the other end of the spectrum, typically paying less — commonly around 5% — in exchange for taking custody, validator selection and tax reporting off the user's hands. JIP-33, Jito's proposed Coinbase integration, would even bridge JitoSOL minting and USDC borrowing into that custodial experience, blurring the line between the models. For a self-custodial holder, the practical framework stays simple: use the LST matching your yield and decentralization preferences, route exits through whichever venue — Jupiter, Sanctum or direct unstake — gives the best realized price for your size, and avoid splitting into so many small LST positions that you cannot continuously track each protocol's risk.

Jito and Marinade remain the two defaults because their scale, audit history, validator governance and public dashboards are the easiest to verify — and because in a stressed market, the venues with the deepest secondary liquidity are where exit spreads stay tight. If that risk ever changes, treat it as a reason to re-evaluate the position immediately rather than waiting for an epoch boundary.

Who should choose which

• Choose JitoSOL to maximize expected yield, if you accept MEV variance and alignment with Jito's client and BAM roadmap — and you track 30-day, not 7-day, APY.
• Choose mSOL for the widest validator distribution, the deepest historical DeFi integration, SOC 2 assurance and the downtime protection built into PSR.
• Choose Marinade Native if you want automated delegation without holding an LST or managing validators yourself.
• Large holders can use directed staking on either system to express validator preferences while keeping a protected floor.
• A common setup: hold both, split toward the trailing-yield leader each quarter, and keep an eye on Jito client concentration as a risk metric.

How to choose in 4 steps

1. Decide what you are optimizing: expected total return favors JitoSOL; decentralization, low variance and compliance breadth favor mSOL.
2. Compare trailing 30-day APY and the MEV component separately; ignore any single-epoch headline.
3. Map your exit: DEX depth for the LST, the delayed-unstake fee, and what instant liquidity would cost in a selloff.
4. If you stack the LST into lending or LP positions, stress-test the mSOL or JitoSOL exchange rate and set a collateral buffer for depeg scenarios.

Frequently asked questions

Sources and further reading

• Jito — official site, JitoSOL and economic hub
• Jito — JitoSOL fee and rewards FAQ
• Marinade — official site and mSOL dashboard
• Marinade — fee model and product documentation
• DeFiLlama — Jito TVL
• DeFiLlama — Marinade TVL