Kamino Finance is the largest lending protocol on Solana with roughly $2.1B in total value locked across its lending markets and automated liquidity vaults as of April 2026. Unlike a pure money market like Aave, Kamino combines lending pools (K-Lend) with automated concentrated-liquidity vaults that auto-rebalance Orca Whirlpool positions. This piece explains how K-Lend works, how the vaults compose with lending, and where the risks concentrate. Stablecoin yield on Kamino has paid 4–9% APY for USDC supply across 2026 depending on borrow demand, making it one of the largest sources of onchain stablecoin yield outside Ethereum.
​

Kamino is a Solana-native money market and concentrated-liquidity automation platform. The protocol launched in 2022 as a Whirlpool-position automator, then expanded into lending in 2023 (K-Lend) and into a multi-product platform by 2025. The combined product is sometimes called a "DeFi super-app" — users can deposit, borrow, leverage, and provide liquidity from a single interface, with the protocol handling the underlying coordination across primitives.
​

Per Kamino's dashboard, the platform's TVL splits roughly 65% into K-Lend (lending pools) and 35% into liquidity vaults that compound Whirlpool positions. The lending pool TVL alone makes Kamino the largest money market on Solana, ahead of MarginFi (~$700M), Save (~$400M), and Drift's spot lending (~$300M).
​

Kamino is governed by KMNO token holders. Voting controls risk parameters, asset listings, and protocol fee distribution. Like most modern lending protocols, the active governance surface is risk parameter changes (LTV ratios, liquidation thresholds, oracle configurations) rather than radical product changes. Solana DeFi apps ranks Kamino against the broader app set.
​

K-Lend is structurally similar to Aave V3 — a pooled lending market where suppliers earn interest from borrowers, with risk parameters set per asset. The mechanism:
​

A user deposits USDC. The deposit mints kUSDC, an interest-bearing receipt token whose redemption value compounds over time as borrowers pay interest. Another user can borrow USDC against eligible collateral (SOL, JitoSOL, mSOL, BTC, ETH, JLP, and others) up to the asset's loan-to-value (LTV) limit. Interest rates float along a kinked utilization curve: low rates when utilization is below a target, sharply higher rates above the kink to discourage full utilization.
​

K-Lend uses Pyth Network oracles to price collateral. Pyth publishes prices on a sub-second cadence sourced from Jane Street, Two Sigma, Jump, and 80+ other publishers. When a borrower's health factor drops below 1.0, anyone can call the protocol's liquidate instruction and seize collateral at a configured discount (typically 5–10% bonus to the liquidator). Liquidation is permissionless and competitive — third-party bots monitor health factors and race to liquidate underwater positions.
​

K-Lend supports isolated markets in addition to the main pool. Isolated markets let the protocol list higher-risk collateral (long-tail tokens, LST derivatives, LP tokens) without risking contagion to the main USDC/SOL/BTC pool. A bug or oracle failure in an isolated market can liquidate users in that market but doesn't cascade to mainline lending positions.
​

The economic design of borrow rates incentivizes specific use cases. Users borrow USDC against SOL collateral to lever long, against JLP to capture stablecoin yield with leverage, or against BTC/ETH to short. Each pattern shows up in observable utilization swings: SOL-collateralized USDC borrows spike during bull rallies; BTC-collateralized USDC borrows show countercyclical patterns.
​

The vaults are the original Kamino product. They automate concentrated-liquidity positions on Orca Whirlpools (Solana's Uniswap-v3 equivalent). A user deposits USDC and SOL into a Kamino vault; the protocol opens a Whirlpool position in a target price range and rebalances automatically when the price moves out of range.
​

Concentrated liquidity offers higher fee yield than constant-product AMMs but requires active management. Without rebalancing, an out-of-range position earns no fees and concentrates 100% in one asset of the pair. Kamino's vault mechanism handles this: the protocol monitors the position, closes it when the range is breached, and reopens centered around the new price. Kamino's documentation details the rebalance triggers and slippage mitigations.
​

Vault fees are paid in the underlying pair tokens, which gives LPs natural exposure to both assets. For USDC/SOL vaults, this means LPs accumulate both USDC and SOL fees over time, with the realized return depending on price path (impermanent loss applies, scaled to range width).
​

Beyond Whirlpools, Kamino has expanded vault offerings to Meteora DLMM pools and other concentrated-liquidity venues. The mechanism is the same: deposit, range-target, auto-rebalance. The differences are pool-specific (Meteora bins behave differently from Whirlpool ranges).
​

For stablecoin teams, Kamino is the largest single source of onchain USDC yield on Solana. K-Lend USDC supply has paid 4–9% APY through 2026, varying with borrow demand. PYUSD listings followed Solana migration; as of April 2026 PYUSD supply on K-Lend pays similar rates to USDC, though pool depth is lower.
​

Three patterns drive observable stablecoin yield:

Stablecoin yield from Kamino is one of several Solana yield options. The full landscape includes stablecoin invoicing platforms for receivables yield, lending markets like MarginFi and Save, and structured products like Drift Spot's specialized pools.
​

Risk parameters define the boundary between solvent and insolvent positions. K-Lend's main parameters per asset:

Liquidation on Solana is mechanically faster than on Ethereum. With 400ms slot times, a position can be flagged liquidatable and liquidated in under a second. This compresses the window for the borrower to top up, but also reduces bad-debt accrual when prices gap.
​

Pyth's price-staleness check is a critical safety: if Pyth hasn't updated a price within a configured threshold (typically 10–60 seconds depending on asset), K-Lend pauses liquidations for that asset. This prevents liquidations on stale prices but freezes risk management during oracle outages.
​

K-Lend's design borrows heavily from Aave V3 but diverges in a few places that matter for users coming from Ethereum:
​

Oracle: Pyth instead of Chainlink. Pyth publishes on a sub-second cadence sourced from first-party financial data publishers; Chainlink publishes on a heartbeat-plus-deviation schedule sourced from third-party node operators. The Pyth model gives Solana lending markets faster price reactions but requires the protocol to handle staleness explicitly. Chainlink's update model is slower but the staleness handling is built into the heartbeat itself.
​

Liquidation runtime. Solana's parallel execution lets liquidations run alongside other transactions in the same slot. Ethereum's serial mempool model creates frontrunning races for the liquidator role. Solana liquidations still race, but the absence of a public mempool changes the competition: liquidators run their own validators or pay priority fees rather than competing in a public auction.
​

SPL token model. K-Lend collateral and borrow tokens are SPL tokens with associated token accounts. Each user-asset combination is a separate ATA. This is administratively heavier than Ethereum's per-contract balance mapping but allows certain optimizations Ethereum lending markets don't have.
​

For users moving between ecosystems, the practical implication is that K-Lend's UX feels similar to Aave (deposit, borrow, repay) but the underlying execution model is distinct. MakerDAO and DAI compares the design space at a higher level.
​

Oracle dependency. Kamino is fully dependent on Pyth for price data. A Pyth manipulation or outage cascades into incorrect liquidations or pause events. Pyth's publisher quorum (median of 80+ publishers) makes manipulation expensive but not impossible. Historical incidents have been short-duration and limited to long-tail assets.
​

Smart-contract risk. Kamino has been audited by OtterSec, Halborn, and Offside Labs (per the protocol's audit page), but composability with Whirlpools, K-Lend, and external venues introduces interaction surface that audits can't fully cover. The protocol has not had a major exploit through 2026, but the structural risk remains.
​

Stablecoin-supply concentration. A large fraction of Kamino's stablecoin TVL comes from a small number of professional treasury and market-maker addresses. Withdrawal of one or two large suppliers can spike utilization and rates abruptly.
​

Vault impermanent loss. Concentrated-liquidity vaults amplify impermanent loss versus constant-product LPs. In sharp directional moves, the vault rebalances after the move, locking in losses. Fee yield offsets this in expectation; path matters.
​

Cross-protocol contagion. JLP is a meaningful share of K-Lend collateral. A failure mode in Jupiter Perps that drives down JLP NAV cascades into K-Lend liquidations of JLP-collateralized positions. This is structural to listing yield-bearing tokens as collateral; the same risk shows up in Aave with stETH/wstETH.
​

Kamino's product surface spans several use cases that overlap with traditional treasury operations:
​

Stablecoin yield. The most common: deposit USDC or PYUSD, earn supply APY. The yield comes from real borrow demand rather than emissions, which makes it more sustainable than reward-token yields. Treasury teams typically allocate a fraction of operational stablecoin float to K-Lend supply, sized against the team's appetite for smart-contract risk.
​

Collateralized borrowing. Treasury teams holding SOL, JitoSOL, or BTC can borrow stablecoins against the position to fund operating expenses without selling. The borrow rate plus liquidation risk is the cost; deferred capital-gains realization is the benefit.
​

Concentrated-liquidity yield. Vaults provide a managed alternative to running Whirlpool positions manually. The trade-off is paying Kamino's vault fee in exchange for not staffing an internal LP-management function.
​

Each use case has corresponding monitoring needs. Treasury teams running stablecoin positions on Kamino at meaningful scale typically integrate alerting on health factor, supply utilization, and Pyth staleness events. Stablecoin treasury APIs covers the broader monitoring stack.
​

Kamino assumes USDC and PYUSD are already on Solana. Getting them there cross-chain is what Eco handles. Eco supports Solana as one of 15 chains and routes stablecoins between Ethereum, Base, Arbitrum, Polygon, and Solana through a unified API. Treasury teams that want to deploy USDC into Kamino's lending pool from an EVM source chain integrate Eco once and get the cross-chain leg handled. The handoff to Kamino is downstream: Eco delivers USDC to a Solana ATA, the team supplies into K-Lend from there. Stablecoin treasury APIs compares the integration patterns; cross-chain liquidity protocols covers the rail set Eco orchestrates across.
​

Kamino has been audited by multiple firms (OtterSec, Halborn, Offside) and has not had a major exploit through 2026. Smart-contract risk and oracle risk remain inherent. Treat Kamino as comparable to Aave on Ethereum: large, audited, but not risk-free.
​

Both are pooled lending markets with Pyth/Chainlink oracles, isolated markets, and permissionless liquidation. Kamino adds automated concentrated-liquidity vaults, which Aave doesn't natively offer. Aave runs on Ethereum and several EVM L2s; Kamino is Solana-only. Solana DeFi apps ranks the broader Solana set.
​

USDC supply APY on K-Lend has ranged from 4% to 9% through 2026, varying with borrow demand. Higher rates correlate with SOL bull moves (more leveraged-long borrowing). Look at the live dashboard for current rates.
​

Yes. PYUSD has been listed on K-Lend since 2025. Pool depth is smaller than USDC, but supply APY tracks similar levels. Digital dollars covers PYUSD issuance specifics.
​

The protocol seizes a portion of your collateral at a 5–10% discount to repay your debt and bring the position back to a safe LTV. The remainder of your collateral stays in your account. Liquidation is partial by default — only enough to restore solvency.
​