Base, Coinbase, Optimism and WalletConnect are trying to make Ethereum wallets feel less like cold storage homework, and EIP-8130 is already live for testing.
Four of crypto's biggest wallet infrastructure names are now pushing the same answer to a problem you've probably felt yourself: seed phrases are a terrible default for normal people. Base, Coinbase, Optimism and WalletConnect are backing EIP-8130, a draft Ethereum standard for native account abstraction that brings passkeys, key rotation, transaction batching and sponsored gas closer to the protocol layer. The test version is live at vibes.base.org, and Base is targeting September for a mainnet rollout through its Cobalt hard fork.
That doesn't happen often. Wallet teams love to talk about better onboarding, but the result is usually another smart wallet feature tucked inside one app, running on one chain. It's built for one growth campaign, then forgotten. EIP-8130 is more ambitious than that. It tries to change what an Ethereum account can do by default.
The cost case is blunt
On Base's test network, USDC transfers built on the new standard cost 63% less than the same transfers run through ERC-4337, the account abstraction standard most smart wallets rely on today. That figure comes from testing data Base published alongside the EIP-8130 specification at eip8130.com. It isn't a tiny optimization. For a payment app or a game, fees that feel small to traders can still feel ridiculous to everyone else. Same goes for a consumer wallet.
The standard was authored by Chris Hunter, an engineer at Coinbase who works on Base, and the official Ethereum Improvement Proposal lists it as Draft status. According to the EIP text, the design introduces a new EIP-2718 transaction type and an onchain Account Configuration system. Accounts can register actors with authenticators, then send transactions that name the authenticator being used, which lets nodes validate the transaction without running arbitrary wallet code first.
No new EVM opcodes. That part matters because it keeps the proposal closer to the machinery chains already understand. Instead of routing transactions through bundlers and a separate entrypoint contract, the way ERC-4337 does, EIP-8130 has the chain validate against account configuration data. The spec's first canonical authenticator set includes k1 signatures, p256 signatures, passkeys and delegate authentication, with room for future algorithms through a companion standards process. That's a more careful claim than saying quantum-safe wallets arrive on day one. They don't.
Seed phrases are the trap. Twelve or twenty-four words, written on paper or, more likely, screenshotted somewhere they shouldn't be, still sit between you and every dollar in your wallet. Lose the phrase and you lose the funds. Hand it to a phishing page and you lose the funds just the same. EIP-8130 tries to make that failure mode optional by letting a wallet authenticate more like your phone already does, through passkeys, scoped keys, recovery flows and permissions a developer can actually design for humans.
The names behind it are the story
Base backing this standard is expected. Coinbase has a direct interest in making wallets easier, and Base needs consumer apps that don't punish users with every click. Optimism and WalletConnect signing on is the stronger signal. Optimism's OP Stack powers Base and other rollups, while WalletConnect is the connection layer that many wallets and dapps already use when they talk to each other.
That's the tell. If EIP-8130 stays a Base-only feature, it may still be useful, but it won't change the wallet experience across Ethereum. If the same pattern spreads across OP Stack chains and gets wallet support through WalletConnect's ecosystem, gasless transactions, social recovery and delegated account control start looking less like add-ons and more like the normal way accounts behave.
There is a bigger use case sitting behind the consumer pitch: non-human wallets. If AI agents are going to transact onchain, they need bounded authority, spend limits, expiry, revocation and a way to act without holding a user's root key. The EIP's actor and scope model is clearly aimed at that world. You don't want an agent with your seed phrase. You want a key that can do one job, for a fixed period, and then stop working.
September is still the hard part
Frankly, the harder test isn't the pitch. It is shipping. EIP-8130 is still Draft in the Ethereum Improvement Proposal process, the same early stage plenty of standards never leave. Base's own upgrade history also argues for caution around the September target. The Block reported that Base delayed its Beryl mainnet upgrade by one day in June after a timing issue with the B20 Activation Registry, and that Cobalt is expected to add native account abstraction alongside more B20 features.
The delay wasn't catastrophic. It was still a reminder. Base is moving faster since it began shipping its own independent upgrades, with Azul in May and Beryl in June, but protocol-level wallet changes touch wallets, apps, nodes and users at once. A test site can prove the flow works. It can't prove that four separate organizations will move in lockstep by September.
The standard works in testing, and the pitch is strong. Now Base has to show that killing the seed phrase is more than a demo people share for a week and forget.
Also read: Across Protocol Got Hacked on Solana and No User Lost a Cent • XRP Withdrawals From Binance Hit Their Highest Level Since 2024 • France Blocks Polymarket Nationwide After Hackers Rigged Weather Bets
