Ethereum and Base Split on Wallet Standards, and Now There Are Two Account Abstraction Camps

Account abstraction was supposed to be the thing that finally made crypto wallets feel like apps instead of tax forms. Instead, it's being built twice, in two different directions, by two groups that have stopped coordinating.

That's the short version of what Derek Chiang, a researcher at Ethlabs, says happened. Talks between Ethereum and Base about a shared wallet standard broke down. No agreement, no joint specification, no timeline for one. Just two roadmaps quietly parting company, with users stuck in the middle wondering why the wallet on their phone behaves like a different product depending on which chain it's pointed at.

Granted, this isn't a chain split. Nobody's forking anything. But it's the kind of quiet divergence that shapes developer priorities for years, and it deserves more attention than it's getting.

What Actually Broke

Some background, because account abstraction has a reputation for being explained badly. The idea is simple enough. Instead of a wallet being a private key that signs transactions, the wallet becomes a small program with its own rules. You can pay gas in stablecoins. You can recover an account if you lose a phone. You can batch five clicks into one signature. That's the promise, and it's a good one.

ERC-4337 was the first serious attempt to deliver it without changing Ethereum's core protocol. It shipped in March 2023 and introduced a whole vocabulary of bundlers, paymasters, and user operations. It works, more or less, but it's bolted on top of the base layer, and bolted-on things are expensive. A user operation on mainnet has historically cost meaningfully more than a plain transfer, which is a hard sell when you're trying to convince someone that this is the easy version.

Then came EIP-7702, which went live with the Pectra upgrade on May 7, 2025. This one lets a regular externally owned account temporarily delegate its code to a smart contract. Same address, same keys, new abilities. Cleaner. Cheaper. Native to the protocol instead of layered beside it.

Base, meanwhile, took a different path. It launched mainnet in August 2023 as an OP Stack rollup under Coinbase, and it's been building wallet infrastructure around 4337-style smart accounts and its own Coinbase Smart Wallet. Cheap blockspace plus a consumer app distribution channel plus a wallet you control is a coherent strategy. Really coherent, actually.

So you had Ethereum leaning toward 7702 for its delegation model and Base leaning toward 4337 smart accounts as the primary path. Those aren't enemies, exactly. They overlap. But every wallet team now has to decide which one they're building for first, and there's no shared answer.

Who Wins, Who Pays

Wallet developers lose. That's the clearest outcome. If you're a small team building a smart wallet and you now have to support 7702 delegation, 4337 user operations, and the Base-specific flavor of both, you've just tripled your surface area for bugs. Infrastructure providers feel it too. Bundlers, paymasters, and indexers all have to model the same user state in more than one way.

Base probably wins in the near term. Coinbase controls the wallet and the chain, which means it can ship a coherent experience without waiting for anyone's consensus. Vertical integration is a real advantage when the horizontal standard is still arguing with itself.

Ethereum wins on principle, if you care about that sort of thing. Keeping the delegation model closer to the base protocol means fewer trusted intermediaries and a smaller attack surface, which matters when the alternative is handing responsibility to third-party bundlers.

Color me skeptical, but I don't think either side is wrong here. Ethereum is protecting its own coherence. Base is protecting its users' experience.

The question worth asking is whether this divergence is permanent or just a phase. Standards in crypto have a habit of fracturing and then quietly re-converging once one approach proves cheaper in production, not in theory.

History suggests otherwise, though, at least on the timescale that matters. Token standards converged because they had to. Wallet standards have more room to stay apart, because the cost of divergence falls on developers and users rather than on the chains themselves.

And admittedly, the numbers are small right now. Smart accounts are still a minority of active addresses across every chain. Base's fees have fallen to fractions of a cent for a simple transfer, which makes the gas savings from either abstraction approach feel less urgent. When transactions are nearly free, the case for rebuild-your-wallet gets weaker.

That's the paradox. Both standards exist to make wallets better. Neither is compelling enough on cost alone to force a winner.

The Takeaway

Expect duplication, and expect it to last longer than anyone's roadmap implies.

If you're building on Base, build for 4337 smart accounts and treat 7702 support as a nice-to-have. If you're building on Ethereum mainnet, do the reverse. If you're building for both, budget the extra engineering time and don't pretend it's free, because it isn't.

The users won't notice the split for a while. They'll notice it the first time they move an account between chains and their recovery setup doesn't come with them. That's the moment this stops being a standards debate and starts being a support ticket.

And here's the part that bugs me. The whole point of account abstraction was to make wallets invisible. Two competing standards make them visible again, in the worst way.

Time will tell, though. Ethereum has a track record of absorbing good ideas from adjacent projects, even when the integration takes years longer than anyone wanted. The skeptics and the proponents are both going to be right about something before this is settled.