Kamino Finance Review 2026

On 13 September 2026, Kamino's main USDC lending market paid depositors 3.51%.

That figure is worth holding onto, because the reviews ranking for this protocol quote 5.08%, a range of 4% to 9%, and a headline of 31.83%. Those numbers are not all describing the same product, and several are readings of the same dashboard taken at different moments. On Kamino the number on the dashboard is not a rate the platform offers you. It is a measurement of how much of the pool somebody else has currently borrowed, recalculated continuously.

That distinction decides whether the yield you see is one you will actually earn. Mixing a leveraged product, a managed vault and a plain deposit into a single "Kamino APY" is how one protocol ends up quoted at both 4% and 31%.

This review works through where Kamino's number comes from, what it did across the last twelve months of data, and the one condition under which it goes highest. That last part matters most, because on a lending market the highest advertised rates are a warning label rather than a bargain, and the clearest statement of that comes from Kamino's own risk documentation.

This is written for someone holding USDC or SOL who has seen a Kamino rate quoted somewhere and wants to know whether it is a number they can plan around.

What Kamino Finance actually is

Kamino is a lending and liquidity protocol on Solana, live since November 2023, holding about $1.3 billion in its lending markets as of 13 September 2026 according to DefiLlama. On the same reading it is the largest lender on Solana, ahead of Jupiter Lend at about $1.1 billion.

It runs four things that matter to a depositor. Lending markets, where you supply an asset and borrowers pay you interest. Lending vaults, where a third-party manager called a curator spreads your deposit across several of those markets. Liquidity vaults, which manage market-making positions on Solana exchanges. And Multiply, which repeatedly borrows and re-deposits to build a larger position than your own capital would allow.

Only the first two are passive in the way a yield seeker usually means. The high figures that circulate in reviews mostly come from the leveraged products, where a price move against you can force your collateral to be sold off to repay the loan, or from vaults whose displayed yield already nets out fees. If the shape of a lending pool is unfamiliar, the mechanics here are the same ones Aave and Compound use, with Solana speed and Solana fees.

One structural point is worth fixing early, because the rest of this comparison depends on it. Kamino is a protocol you use directly from your own wallet, which stays in your control throughout. EarnPark, further down, is an account with a company that deploys the deposit on your behalf. Those are different arrangements with different failure modes, and no rate comparison between them is a like-for-like swap.

Kamino has a governance token, KMNO, launched 30 April 2024 with a total supply of 10 billion, per its token documentation. It does not gate the lending rate. Neither platform in this comparison hides its published rate behind a token, so that axis simply does not separate them.

Where the number comes from

Kamino sets rates with a utilization curve. Utilization is the share of a pool that is currently lent out. Kamino's own fee documentation puts the depositor side plainly, that supply APY is approximately the borrow rate multiplied by utilization.

Two consequences follow, and both matter more than any headline figure.

The first is that idle money dilutes you. If borrowers pay 8% and utilization sits at 70%, lenders earn roughly 5.6%, because the other 30% of the pool earns nothing and is averaged into everyone's return. Kamino describes that spread as the structural cost of instant liquidity.

The second is that the protocol takes a cut before you see anything. Kamino publishes the spread per asset per market. USDC in the main SOL/BTC market carries a 15% protocol spread, USDC in the JLP and Altcoins markets 20%, and SOL 11%. As its documentation notes, a 15% spread on a 10% borrow rate means lenders receive 8.5%, not 10%.

The curve itself is deliberately steep. Kamino builds in a sharp bend at a target utilization level, described in its interest rate model. Below target, rates rise gently. Above it, they rise violently. The documented worked example has a stablecoin pool at 85% utilization producing a 53.3% borrow rate and a 38.5% supply rate. Target utilization for stablecoins is set at 80% to 90%.

So a very high Kamino rate is not a generous offer. It is the protocol shouting that a pool is nearly drained and paying whatever it takes to attract deposits and force repayment.

What the rate actually did over twelve months

Readings below are from the DefiLlama pool history for Kamino's main USDC market, 366 daily datapoints between 13 September 2025 and 13 September 2026. Every reading carries a zero token-reward component, so this is lending yield alone, not a figure propped up by handing out free tokens.

The second column needs a word of explanation, because it is the one that matters. It is not the size of the market. It is the portion of it sitting unborrowed, which is the money actually available for lenders to withdraw. On 13 September 2026 the market held $124.5 million of deposits against $101.2 million of loans, leaving $23.3 million available.

The median across the full year is 3.69%. The rate sat below 4% on 225 of the 366 days and below 2% on 36 of them. It reached 8% or higher on six days and 10% or higher on three. So the realistic expectation for this market is a little under 4%, and the double-digit readings are rare.

Now read the two columns against each other, because the relationship between them is not a coincidence. It is the formula. Supply APY rises with utilization, and utilization is high exactly when little is left unborrowed. The worst rate of the year, 1.13%, came on a day with $127.2 million available. The best rate of the year, 11.23%, came on a day with $4.8 million available. Sorted across all 366 days, readings with under $10 million available ran a median of 5.10%, while readings with over $80 million available ran a median of 2.70%.

That is not a flaw in Kamino, and the same arithmetic governs every utilization-priced venue, including Compound and Aave. But it has a blunt practical meaning that no rate card shows you. A high advertised rate on a lending market is a statement about how little of that market is currently free to leave. A depositor who moved money in on the strength of 11.23% in late August was earning 3.5% within a fortnight.

The condition under which the rate goes highest

Here is the part that changes how the number should be read, and it is Kamino's own wording rather than an outside critique. From its risk framework, on liquidity risk, when borrowing demand exhausts the supply in a pool and utilization reaches 100%, "lenders cannot withdraw until borrowers repay or new supply arrives. This does not cause financial loss, in fact, it produces exceptionally high interest rates for lenders, but it temporarily restricts withdrawals."

Read that twice. The highest rates a lending market can display and the inability to get your money out are not two separate risks. They are the same event, described from two sides, and the table above is that statement in numbers. Kamino's interest rate documentation is equally direct, noting that lenders who need their funds are locked out, and that new deposits are discouraged because, in its own phrasing, why deposit if you might get locked.

There is a real episode in the data. On 18 April 2026 an attacker drained 116,500 rsETH, about $292 million, from Kelp DAO's bridge, the system that moves its token between blockchains, in what CoinDesk called the biggest crypto exploit of the year. Emergency freezes followed across Aave, SparkLend, Fluid and Upshift.

Kamino was not one of the protocols named. It was not exploited and it recorded no bad debt, meaning no loans went unrecovered in a way that would push losses onto depositors. Yet in the same DefiLlama series, its main USDC market went from 3.00% on 18 April to 10.37% on 19 April, while the money available to withdraw fell from $20.6 million to $6.8 million.

The public data does not say whether that was lenders heading for the exit, borrowers drawing more, or both, so it would be wrong to claim one. For a would-be depositor the distinction changes nothing. A yield seeker scanning Solana rates on 19 April, the day after the largest exploit of the year, would have seen one of the highest Kamino USDC readings in months, and that reading was high because the withdrawal buffer had thinned to a third of the previous day's. The rate and the reason to be careful were the same number. That is why market liquidity deserves more attention than the rate printed next to it.

Kamino does build against this. Its curve is calibrated to make full utilization economically unsustainable, it applies daily caps on how fast borrows accumulate, and its lending vaults hold a 5% to 10% unallocated buffer so small withdrawals clear instantly regardless of conditions. Its documentation is candid that in market-wide stress, or under a deliberate attack where someone borrows the entire supply and simply pays the high interest to hold it, the curve alone may not be enough.

What Kamino gets right

A review that only pressed on the rate would be misleading, because Kamino's engineering and disclosure record is strong and better than most of what this sector publishes.

Its security page lists 20 external audits across five firms with zero critical vulnerabilities found, plus four completed formal verifications from Certora and OtterSec, the most recent covering Kamino Lend in March 2026. Formal verification means proving mathematically that specific properties of the code always hold. Kamino's own framing is that auditing is probabilistic while verification is mathematical, and whatever weight you give that distinction, few platforms in this space commission it at all.

The operating record supports the code. Kamino reports zero bad debt since November 2023 across five named stress events, each listed with dates and collateral seized. The most severe was 5 to 6 February 2026, when SOL fell 18% over 48 hours, producing 55,649 liquidation events across 30,030 wallets and $19.36 million in collateral seized, with no bad debt generated. Those are Kamino's own figures rather than audited ones, but they are published with enough specificity to be checked on-chain, which is more than most disclosures allow.

Since early 2025 the risk consultancy Allez Labs has published monthly risk reports to Kamino's public governance forum, more than 14 to date, covering utilization trends, vault performance and stress scenarios modelled to a 60% price shock. Vault curators can also lock their own capital in an insurance pool they cannot exit quickly, and vaults can be permanently restricted to protocol-approved pools so a compromised curator key cannot route deposits into a fraudulent market.

Those are genuine strengths. Nothing in this review is an argument that Kamino is badly built.

The costs and risks reviews skip

Kamino charges no deposit fee, and Solana transaction costs are negligible. The costs sit elsewhere.

The protocol spread described earlier is the main one and it is invisible on the rate card, because the number you see is already net of it. Lending vaults add a performance fee on profits and an annual fee on assets under management, both deducted before the displayed APY, along with a withdrawal penalty which, unusually, is returned to the vault rather than paid to the curator.

The risk most depositors never price is curator discretion. Kamino's vault risk page states that curators change allocations without any governance approval, and that depositors bear the risk of every allocation decision made after their deposit. Choosing a vault is choosing a manager, and the protocol says so directly. If a borrower defaults and the loan cannot be recovered, lenders including the vault absorb the losses in proportion to their share.

None of this is hidden. It is all in the documentation, which is to Kamino's credit. It is simply not in the reviews, and it is not in the APY.

Who Kamino is best for

Kamino suits someone already operating on Solana with a wallet, who is comfortable reading a utilization figure before depositing rather than a headline, and who treats the rate as a live market price to be monitored rather than a yield to be set and forgotten. For that user it is among the better engineered options on the chain, and the Solana DeFi landscape has thinner alternatives.

It suits borrowers and leveraged users better still, since that is where the protocol's design effort concentrates.

It fits poorly if you want a rate you can plan around, if you would not notice your yield falling from 11% to 3.5% over two weeks, or if you would read a sudden rate spike as good news. It also fits poorly if you have no interest in evaluating yield farming risks or curator track records before committing capital.

Kamino compared with EarnPark

This is a rates and terms snapshot, not a safety ranking. The two run different models, and the first row matters as much as the rate rows.

The same question this review asked of Kamino should be asked here, which is where the money comes from. Kamino's answer is borrowers paying interest on overcollateralised loans, net of the protocol spread. EarnPark's is a trading desk rather than a lending book, and the strategies behind the USDC rate are named. Maker Core is market making, quoting both sides of an order book and earning the spread. DeFi is automated concentrated liquidity. Both earn from trading volume rather than from prices rising, which is set out on the page covering how EarnPark makes money, alongside the years individual strategies lost money.

Both sides of that table need their caveats stated rather than buried.

On Kamino, 3.51% is what the main market paid on the reading date and it will be different by the time you read this. On EarnPark, the published USDC rate of up to 10% as of September 2026 is the base any user receives holding zero PARK, with the PARK token boost sitting on top as optional upside for holders rather than the route to the advertised figure. EarnPark's rates move too, and a published rate is not a promise about next quarter. What it is not is a live readout of somebody else's borrowing.

Withdrawal terms differ by strategy and deserve reading before the rate does. As of September 2026, USDC Maker Core carries a 30 day bonding period with a one day cancellation window, during which yield continues accruing in full, and a $2 flat fee. USDC DeFi withdraws instantly for 0.30% plus a $3.20 flat fee. The top of EarnPark's wider range, the 15% on USDT and 20% on SOL, comes from high risk Alpha Vault strategies that settle monthly with a 24 hour freeze after opening a position, which is a materially different commitment from either of the USDC options above.

For anyone comparing across chains rather than within Solana, the stablecoin yield landscape and the split between CeFi and DeFi models are worth reading alongside this.

On SOL specifically, Kamino's main market paid 4.52% on 13 September 2026. EarnPark's SOL page publishes up to 20%, again from a high risk monthly-settling strategy, with low risk SOL Maker Core at 7% and medium risk SOL DeFi at 6.5% as of September 2026. Anyone weighing SOL yield options should also look at how staking SOL compares, on a risk-adjusted basis rather than by raw rate. EarnPark's own risk statement sets out what the strategy tiers mean.

Verdict on Kamino vs EarnPark

Kamino is well built, unusually well documented, and has a clean operating record across five stress events. The problem is not quality. It is that its number answers a different question from the one most yield seekers are asking.

Kamino tells you what a pool is paying right now given how much of it is borrowed. That answer is honest, live, and unstable by design, and its most attractive readings coincide with its least comfortable conditions. A little under 4% is the realistic expectation from the main USDC market on a year of evidence, not the 11.23% peak or the double-digit figures from leveraged products quoted elsewhere.

EarnPark answers the other question, which is what rate you are being offered before you commit. Up to 10% on USDC as of September 2026, published per strategy with the risk tier and the withdrawal terms stated next to it, available to any user holding no PARK. Read those terms first. The 10% low risk route carries a 30 day bonding period, and the highest headline figures across the platform come from high risk strategies that settle monthly.

If you want to watch a market, Kamino is a good one to watch. If you want a rate you can plan around, the comparison is not close. Compare the terms alongside Solana lending alternatives, decide which question you are actually asking, and then Start earning.

Sources used

• Kamino, risk framework and track record, https://kamino.com/docs/risk
• Kamino, interest rate model, https://kamino.com/docs/risk/safeguards/interest-rates
• Kamino, fees and interest rates, https://kamino.com/docs/products/borrow/fees
• Kamino, lending vault risks, https://kamino.com/docs/products/lending-vaults/risks
• Kamino, curators and fees, https://kamino.com/docs/products/lending-vaults/curators
• Kamino, KMNO token, https://kamino.com/docs/kmno
• Kamino, security, audits and formal verification, https://kamino.com/security
• DefiLlama, Kamino Lend protocol TVL, https://defillama.com/protocol/kamino-lend
• DefiLlama, Kamino main USDC market pool history, https://defillama.com/yields/pool/d2141a59-c199-4be7-8d4b-c8223954836b
• CoinDesk, Kelp DAO exploited for $292 million, 18 April 2026, https://www.coindesk.com/tech/2026/04/19/2026-s-biggest-crypto-exploit-kelp-dao-hit-for-usd292-million-with-wrapped-ether-stranded-across-20-chains
• EarnPark live asset pages and help centre, read 13 September 2026