How Aave interest rates work

On Aave, nobody sits in an office deciding what borrowers pay. Every asset in every market has its own pool, and a formula sets the rate from one input: how much of the pool is currently borrowed. When a lot is borrowed, rates rise. When the pool is mostly idle, rates fall.

Utilization: the number that drives the rate

Utilization is the share of supplied funds that is out on loan:

utilization = total borrowed / total supplied

If suppliers have deposited 100 million USDC and borrowers have taken 80 million, utilization is 80%. The remaining 20% is spare liquidity that suppliers can withdraw at any time.

High utilization is efficient, but it is risky for suppliers: if nearly everything is lent out, they may have to wait to withdraw. The rate formula is built to stop that from lasting.

The kinked rate curve

Aave V3 uses a curve made of two straight lines that meet at a point called the optimal utilization. Four parameters describe it:

• Base rate: the borrow rate when nothing is borrowed.
• Slope 1: how much the rate rises between 0% utilization and the optimal point.
• Optimal utilization: the kink, where the gentle line ends and the steep one begins.
• Slope 2: how much the rate rises between the optimal point and 100% utilization.

Below the kink:

rate = base + slope 1 x (utilization / optimal)

Above the kink:

rate = base + slope 1 + slope 2 x ((utilization - optimal) / (1 - optimal))

At the time of writing (September 2026), USDC on the Ethereum Core market used a base rate of 0%, a slope 1 of 4.4%, an optimal utilization of 94% and a slope 2 of 10%. With those settings the borrow rate works out like this:

These are APRs. The Aave app and this site display rates as APY, which is slightly higher. See APR vs APY on Aave.

Why rates jump above the kink

Look at the last three rows. Going from 0% to 94% utilization adds 4.4 points of interest. Going from 94% to 100% adds another 10 points. The steep part is deliberate. Once a pool is nearly empty, a sharply higher rate does two things at once: it pushes borrowers to repay, and it attracts new suppliers who want the higher yield. Both bring utilization back down, which keeps liquidity free for withdrawals.

It also explains sudden spikes. If one large withdrawal or loan pushes utilization from 93% to 100%, the rate can triple within a single block and fall back minutes later when the market reacts. Those short spikes still cost borrowers money, as we show in why your average rate differs from the current rate.

Where the supply rate comes from

Suppliers are paid out of the interest that borrowers pay. Two things make the supply rate lower than the borrow rate. First, interest from the borrowed part is shared across all supplied funds, including the idle part. Second, the protocol keeps a cut called the reserve factor, which goes to the Aave treasury.

supply rate = borrow rate x utilization x (1 - reserve factor)

Using the USDC example, where the reserve factor was 10%: a 4.34% borrow rate at 92.7% utilization gives 4.34% x 0.927 x 0.90 = 3.62% for suppliers.

Who sets the parameters

Aave governance chooses the four curve parameters and the reserve factor separately for each asset in each market, and changes them by vote as conditions change. Stablecoins often get a gentle first slope and a high optimal utilization. More volatile assets often get a lower optimal point, so that more liquidity stays free. The full specification is in the Aave documentation.

You can see the live settings, and where utilization currently sits on the curve, on any asset page, for example USDC on Ethereum. To compare assets, start from all markets.