Aave Delegate Platform Proposes Pausing Three L2 Deployments Citing Weak Revenue

A governance delegation platform for Aave, the largest decentralized lending platform, with more than $29 billion in total value locked (TVL), has proposed pausing three underused Layer 2 deployments of Aave V3.

In a Jan. 29 governance proposal that moved to a snapshot vote on Feb. 3, the Aave Chan Initiative (ACI) proposed that Aave freeze its V3 deployments on Ethereum L2s zkSync Era, Metis, and Soneium to cut costs.

“Over time, it has become clear that a small subset of instances contributes very little user activity, TVL, and revenue, while still requiring a non-trivial amount of attention from service providers and governance participants,” ACI wrote in the prospal.

The proposed reduction in L2 deployments aims “to reduce operational overhead and governance burden by addressing instances that are clearly non viable today.”

Among the three networks, zkSync currently has the largest TVL at about $26 million, followed by Soneium with $21.6 million and Metis with $11.7 million, according to DefiLlama data.

Over the past 30 days, Aave generated just $714 in revenue on zkSync, $679 on Metis, and just $150 on Soneium, per DefiLlama. For comparison, within the same timeframe Aave made over $7.7 million on Ethereum and nearly $298,000 on Base.

Now, ACI is pushing for stricter terms on future expansions. The proposal calls for any new chain deployment to guarantee Aave a minimum of $2 million in annual revenue, arguing that the protocol’s liquidity is often underpriced given the “upfront and recurring costs.”

The snapshot vote on the proposal, which runs through Feb. 7, has so far drawn unanimous support, with 257,300 votes in favor and none against.

Voting kicked off the same day that Ethereum’s broader scaling strategy came under renewed scrutiny. As The Defiant reported earlier this week, Ethereum co-founder Vitalik Buterin published an X post arguing that the rollup-centric roadmap for the network “no longer makes sense,” and arguing that L2s should focus on other use cases.