Solana staking pools are smart contracts that gather delegated SOL, spread it across a validator set, and issue a liquid staking token that tracks the position. A pool creates no yield of its own — it collects yield that already exists on-chain and passes it through, minus fees. That yield has exactly three sources: inflationary SOL emissions, the validator’s share of transaction fees, and MEV tips paid by traders for block ordering. In August 2026, emissions supply roughly nine-tenths of the total, and a delegator nets around 6% APY, with MEV-heavy pools like Jito pushing closer to 7%.
Key Facts
- About 68% of SOL supply is staked in 2026 (~425-432M SOL) — the biggest single input into every pool’s APY.
- Network inflation sits near 3.6-3.7%, falling 15% a year toward a 1.5% terminal floor.
- Base emission yield ≈ inflation ÷ staked share: 3.6% ÷ 0.68 ≈ 5.3% gross, before commission.
- MEV tips add ~0.5-1%. The Jito-Solana client runs under 95%+ of active stake, and Jito tips make up over 60% of priority-fee volume.
- Largest pools by TVL: Jito ~$1.4B, Sanctum ~$1B, Marinade ~$740M, BlazeStake ~$200M.
- Pool fees stack on validator commission — Jito takes 5% of tips, Marinade charges ~0.16% annual management plus 0.3% on withdrawal.
- SIMD-0550, proposed June 2026, would double the disinflation rate to 30% per year and cut the path to the 1.5% floor from ~5.7 years to ~2.8.
Source One: Inflationary Emissions
Every epoch, Solana mints new SOL on a fixed schedule and hands all of it to stakers, minus validator commission. Emission started at 8% in 2021 and drops 15% a year, putting 2026 near 3.6%. This is the load-bearing component of staking APY: strip it out and pool yields collapse under 1%. It also makes staking closer to dilution defence than income — non-stakers fund it through a shrinking share of supply, as detailed in our breakdown of Solana’s supply and burn model. The delegation walkthrough itself lives in the guide to staking SOL from a wallet; Gem Wallet, Phantom, and Solflare all handle native stake and pool tokens.
Why the Staked Ratio Sets Your APY
Emissions are a fixed pie split among however much SOL is staked. Divide inflation by the staked share and you get gross emission yield: 3.6% ÷ 68% ≈ 5.3%. That single equation explains most APY movement traders blame on validators — when participation climbed through Q2 2026, the network staking rate drifted from ~5.9% to ~5.5%, because the same emissions got split more ways. The corollary is uncomfortable for pools: as inflation steps down 15% a year while participation stays high, the emission component compresses no matter how good a pool’s validator selection is.
Source Two: Transaction Fees
Solana’s base fee is 5,000 lamports per signature, half burned and half paid to the block leader. Priority fees — what traders bid to jump the queue during congestion — go entirely to the validator producing the block. This is the thinnest of the three slices, usually a fraction of a percentage point of APY, because base fees are trivially cheap by design. It only becomes visible during heavy memecoin activity, when blockspace demand spikes for a handful of epochs. Per-epoch stake account rewards are inspectable on Solscan.
Source Three: MEV Tips
MEV tips sit outside the protocol entirely. Validators running the Jito-Solana client auction the right to place transaction bundles at the top of a block; searchers arbitraging DEX price gaps pay tips to win those slots, and the tips flow back to stake accounts net of MEV commission. This is where pool APYs actually diverge — a pool delegating only to MEV-enabled validators captures 0.5-1% that a pool ignoring MEV never sees. Tips also scale with volatility rather than inflation, which is why MEV-weighted yields swing hard through memecoin weeks while emission yield barely moves.
What the Pool Adds and What It Takes
A pool’s job is delegation strategy plus tokenisation: it picks the validator set, rebalances between them, and mints an LST that can be sold, lent, or posted as DeFi collateral without unstaking. For that it charges a second fee layer above validator commission. Pool sizes are trackable on DefiLlama.
| Pool | Token | Validator Strategy | Fee Model |
|---|---|---|---|
| Jito (~$1.4B) | JitoSOL | Only Jito-Solana client validators | 5% of MEV tips |
| Marinade (~$740M) | mSOL | 400+ validators, excludes top-32 group | ~0.16%/yr + 0.3% withdrawal |
| BlazeStake (~$200M) | bSOL | 200+ validators, widest set on Solana | Fee on staking rewards |
| JupSOL | JupSOL | Jupiter-operated validator | 5% of base rewards, 0% deposit |
Two pools with identical inputs still report different APY, and the gap traces to four things: MEV capture, average commission of the validator set, the pool’s own fee, and how fast it rotates stake off underperformers.
Pool Versus Direct Delegation
Delegating to one validator gives you that validator’s exact performance — no smart contract sits between you and your stake, and commission is the only cost. The trade-off is concentration: if that node misses vote credits for a week, you eat the full shortfall. A pool averages performance across hundreds of validators, so one node going dark barely dents the rate, and the LST removes the exit queue. You pay for that in contract risk and a second fee layer. Direct delegation is the conservative structure; pools are the liquidity structure.
Where the Yield Actually Breaks
Solana still has no protocol-level slashing, so validator misbehaviour costs rewards rather than principal — a milder risk profile than most proof-of-stake chains. The erosion points are fee stacking (validator commission plus pool fee plus MEV commission on the same yield), downtime, LST depeg during forced selling, and contract failure in the pool itself. The structural risk is separate: emissions are scheduled to shrink, and SIMD-0550 would accelerate that sharply. An earlier attempt, SIMD-0228, was rejected in 2025 with 61.4% support against a 66.7% threshold — but every proposal that does pass lowers the base component of every pool’s APY at once.
Fixed Supply as the Other Side of the Trade: $DOLAN
Staking yield exists because SOL supply expands — stakers are simply the ones who get the new tokens instead of being diluted by them. A fair-launch memecoin runs the opposite structure: DOLAN Duck ($DOLAN) has a fixed 98.3M supply on Solana with no emission schedule at all, so there is no dilution to defend against and no yield to defend with. Holder count, roughly 10,700, is the variable that moves instead of supply. Funding a speculative sleeve from pool income pairs a slowly compressing yield stream against an asset with zero issuance — and exit timing is the underestimated part, since LSTs sell instantly while native stake sits in a cooldown, as covered in how long unstaking SOL takes.
From three on-chain sources: inflationary SOL emissions (the large majority, ~3.6% network inflation in 2026), the validator’s share of transaction and priority fees (a fraction of a percent), and MEV tips paid by searchers for block ordering (roughly 0.5-1%). The pool itself generates nothing — it routes and tokenises existing yield.
Four variables: whether the pool captures MEV tips, the average commission of its validator set, the pool’s own fee, and how quickly it rotates stake away from underperforming validators. Emission yield is identical for everyone, so all visible APY differences come from those layers.
Not automatically. A pool spreads risk over hundreds of validators and gives you a tradeable token instead of an exit queue, but it adds a smart contract and a second fee layer. Direct delegation is cheaper and has no contract risk, at the cost of depending on one node’s uptime.
Not through validator behaviour — Solana has no protocol-level slashing today, so poor performance costs rewards rather than principal. The genuine loss vectors are a bug or exploit in the pool contract and an LST trading below its underlying value when you need to exit fast.
Roughly 0.5-1 percentage point in 2026, depending on network activity. Tips scale with trading volatility rather than with the emission schedule, so MEV becomes a larger share of total yield in busy weeks and during memecoin cycles.
Yes. Inflation falls 15% per year toward a 1.5% terminal rate, which drags the emission component down for every pool at the same time. SIMD-0550, proposed in June 2026, would double that disinflation rate to 30% annually and compress the path to the floor from about 5.7 years to 2.8.
Yes, and the layers compound. Validator commission is taken first, then the pool fee — Jito takes 5% of MEV tips, Marinade charges around 0.16% annually plus 0.3% on withdrawal, and some pools take a percentage of base staking rewards. Always read the fee model, not just the headline APY.