Aave v4 just went live on Ethereum. This is the first major overhaul since v3 and introduces a new hub‑and‑spoke architecture designed to make lending more modular, expand into real‑world credit markets, and improve risk management.
The launch, however, comes at a turbulent moment for Aave’s governance. Key independent contributors such as Bored Ghosts Developing (BGD Labs) and the Aave Chan Initiative (ACI) have announced they are winding down their work after clashing with Aave Labs over budget and control as outlined in our memo.
Below is a quick overview of what v4 does, which projects are building on it, and how governance tensions are unfolding amidst the launch.

What v4 changes
Earlier versions of Aave pooled all deposits into a single market with unified risk parameters. V4 replaces that with a modular hub-and-spoke model where a central liquidity Hub holds the capital, while Spokes are separate lending markets that can draw credit lines from the hub and set their own collateral rules, borrowing parameters and risk limits. This allows different risk profiles such as institutional borrowers, tokenised asset markets or strategy‑specific vaults to share the same liquidity base without forcing all users into the same pool.
The initial deployment launches three distinct hubs:
Core is the general‑purpose hub for most borrowers.
Prime targets suppliers who want to use ETH and BTC as collateral without being exposed to variable‑rate borrower demand.
Plus is tailored for Ethena‑adjacent strategies. Each hub can issue credit lines to multiple spokes, and the DAO can create new hubs as needed.
V4 also introduces a tiered interest‑rate mechanism where borrowers pay a base rate plus a user‑specific risk premium depending on the spoke they interact with. A new liquidation engine adds a target health factor, a variable liquidation bonus, and dust prevention, aiming to reduce liquidation cascades and eliminate tiny, uneconomical positions. These changes are meant to improve capital efficiency and risk management across the varied spokes.
Supported protocols at launch
At launch, a few dedicated spokes have already went live including: Lido, EtherFi, Kelp, Ethena, and Lombard. Each spoke represents a separate market with its own collateral and borrowing rules, drawing liquidity from the hubs. For example:
Lido allows borrowing against stETH and other liquid‑staked ETH positions.
EtherFi targets EigenLayer‑secured liquid staking collateral.
Kelp offers structured yield strategies built on stETH.
Ethena focuses on synthetic dollar markets, using USDe and other derivatives.
Lombard is designed for tokenised real‑world assets, including gold (XAUT) and tokenised treasuries.
Supported assets across v4 include USDT, XAUT, USDC, EURC, cbBTC, frxUSD and USDG, with Chainlink serving as the exclusive oracle provider. Aave Labs also launched Aave Pro, a new interface built for advanced users and institutions, and signalled that multi‑chain deployments (such as on Avalanche) could follow pending DAO approval.
Governance turbulence: BGD and ACI Exit
The v4 launch coincides with the fairly recent governance dispute. BGD Labs, an independent development team that built much of Aave’s v3 codebase, announced it would stop working with the Aave DAO when its contract expires in April 2026.
In a blog post, BGD said it was increasingly pressured to prioritise v4 over the more mature v3 and disagreed with Aave Labs’ “aggressive” push to migrate users to the new system. BGD noted that v3 remains stable and that focusing entirely on v4 could be risky.
Shortly afterward, Marc Zeller’s Aave Chan Initiative (ACI), responsible for roughly 61% of governance actions over 3 years, said it would wind down operations. Zeller cited a governance vote that approved a record budget for Aave Labs and alleged that the vote was driven by addresses linked to Aave Labs rather than the broader community. He argued that there was “no role for an independent service provider” when the largest budget recipient used undisclosed voting power to pass its own proposal.
The departures have raised questions about how Aave will manage its technical roadmap and governance going forward. Stani has said the protocol will continue to operate normally and that the DAO can still create and manage spokes. However, it remains to be tested whether the new modular v4 design and new credit markets can attract the institutional liquidity Aave hopes for while navigating these governance headwinds.