The most common Ethereum failure is not a hack. It is a wallet full of USDC that cannot move because there is no ETH left for gas.

Core developers have now put a protocol fix on the calendar. On the August 27 All Core Developers Execution call (ACDE #244), EIP-8141 — Frame Transactions — moved from “Considered for Inclusion” to Scheduled for Inclusion in Hegotá, the upgrade aimed at 2027. That is not mainnet. It is a locked engineering slot. The Ethereum Foundation Protocol Cluster later ranked Frames an S-tier must-ship for Hegotá, alongside consensus-layer FOCIL (EIP-7805). Vitalik Buterin, one of ten listed authors, said in early September that quiet work on the spec had moved faster than the public thread suggested.

What Frames actually change

Today a transaction is one blob: one sender, one signature scheme, one gas payer, and that payer must be ETH. Frame Transactions split that blob into a sequence of ordinary contract calls — up to 64 frames inside one atomic type-0x06 envelope. One frame can validate. One can approve who pays gas. The rest execute. The account that starts the action and the account that settles the fee no longer have to be the same.

That is the UX punchline. A payments app can sponsor the first swap for a user who holds only a stablecoin. A paymaster can take USDC in one frame, sell it for ETH in the next, and hand validators the ETH the protocol still demands. The user never needed an ETH balance. The chain still gets paid in ETH. That distinction matters. This is not “USDC gas at the consensus layer.” It is “ETH gas, with the conversion and sponsorship expressed on-chain in the same transaction.”

Draft constants put intrinsic cost at 12,000 gas plus 475 per extra frame. Early January text was looser (15,000 intrinsic, a 1,000-frame cap). The numbers have already tightened. They will move again. Treat them as direction, not a fee schedule.

Why this is not just another 4337 wrapper

Ethereum has been talking about account abstraction since EIP-86 in 2016. ERC-4337 and EIP-7702 (Pectra, May 2025) got wallets part of the way: smart accounts, temporary delegation, off-chain bundlers. Those systems work. They also fragment. Different mempools, different paymasters, different rules if the bundler goes dark.

Frames are meant to pull that logic into the public mempool and the EVM. Existing externally owned accounts are in scope — users are not forced onto a new contract wallet to get sponsored gas. Validation can be a contract call instead of a hardcoded secp256k1 check. That is also the quantum door: rotate an account to a post-quantum scheme without moving funds to a new address, and without a dedicated hard fork per algorithm. Hegotá’s other job, in the Foundation’s framing, is to make that key-rotation path real enough to aim at a quantum-resistant L1 later in the decade.

Buterin’s longer write-up in February described a RAI-for-gas path as validation, paymaster check, token transfer, execution, refund — no off-chain relayer required if the chain itself is up. Intermediary minimization is the political line. Whether wallets and apps actually ship that flow, or keep using the same relayers with a prettier envelope, is an adoption question Hegotá cannot answer from a spec.

What is not decided

Scheduled for Inclusion is not frozen code. On the August call, developers SFI’d Frames in the last minutes with an explicit caveat: the EIP number, the implementation, and almost everything in the draft can still change. A competing design, EIP-8130, is still in the conversation. Glamsterdam — the upgrade before Hegotá, targeted for later 2026 — has to land cleanly first. Sepolia and Hoodi dates for Frames are blank. Public-mempool admission will likely start conservative because new validation paths are a denial-of-service surface. ERC-7562-style restrictions are already in the trade-off list.

So the honest sentence is narrower than the headlines. Ethereum has committed to trying to let the sender and the gas payer diverge at the protocol layer in 2027, while keeping ETH as the asset validators receive. If that ships, the “buy ETH before you can spend USDC” tax on new users gets a real off-ramp. If the spec mutates into another bundler-shaped thing, L2s and 4337 keep doing the UX work they already do.

For now the news is process, not product. Hegotá has a headliner on the execution layer. The empty-ETH wallet is the problem it is supposed to kill. The test is a mainnet transaction, years out, where the only balance in the sending account is a stablecoin — and it still goes through.