If you use onchain apps, you have probably run into this: a new wallet receives stablecoins, but you cannot send them because you have no ETH for gas. You try to swap tokens on a DEX and get stuck. Or you withdraw USDC from an exchange, only to realize you forgot to set aside ETH for the next transaction.
Then comes the familiar scramble to buy a little ETH or ask someone to send you enough to get started. It is one of crypto’s most reliable ways to turn away newcomers.
That is why Vitalik Buterin’s recent comments on the progress of EIP-8141, or Frame Transactions, have attracted attention. If the proposal is adopted, a user could transfer assets without holding ETH in their wallet. The network fee could be shown as a USDC charge, or an app could cover it entirely.
The news quickly produced an alarming claim: “If you don't need ETH to use Ethereum, what value will ETH have?”
The answer is not that simple. What a user pays with and what the Ethereum protocol ultimately uses to settle transaction fees are two different things.
1. What Does “Paying Gas Without ETH” Actually Mean?
Consider an ordinary Ethereum transaction today.
Frank has 1,000 USDC in his wallet and wants to send 100 USDC to a friend. If he has no ETH, he cannot submit the transaction.
That is because, for a standard Ethereum account today, the person initiating a transaction also pays its network fee. Frank signs, Frank sends, and Frank’s ETH balance pays for gas. Those roles are effectively bundled together.
EIP-8141 proposes separating them.
Under the current proposal, a Frame Transaction can contain multiple frames that handle different parts of the process, including transaction verification, determining who pays gas, and execution of the user's intended action.
In Frank’s case, he could still sign to confirm, “Yes, I want to send this 100 USDC.” But a separate sponsor or paymaster could provide the ETH needed to pay Ethereum’s gas fee. Frank could then pay that sponsor a small amount of USDC.
Frank might see only “Send 100 USDC” and “Network fee: 0.1 USDC.” He would not need to know the ETH-denominated gas cost or prepare ETH in advance.
From the protocol’s perspective, however, Ethereum has not started accepting USDC as gas. Someone else paid the network fee in ETH and settled separately with Frank in USDC.
EIP-8141’s own example of paying fees with an ERC-20 token follows a similar structure: a sponsor approves and covers the gas payment, while the user transfers tokens to the sponsor as compensation.
A familiar payment analogy may help. If you use a card billed in your home currency while shopping in Japan, you see a charge in your own currency, while the merchant may receive yen. That does not mean Japan has changed its settlement currency. The asset shown at the point of payment and the asset used for settlement can differ.
EIP-8141 offers a similar abstraction. The user may no longer need to hold ETH personally: another account pays the ETH-denominated gas fee, then recovers the cost in USDC or another asset.
That distinction sounds small, but it leads to a very different conclusion about ETH.
2. Why Does Ethereum Want to Hide Gas from Users?
Not having to prepare ETH yourself is not an entirely new idea.
Account abstraction under ERC-4337 already allows smart accounts to use paymasters for gas sponsorship. An app can cover a user’s fee, or let the user pay the fee in stablecoins.
The difference is where the machinery sits. ERC-4337 adds infrastructure outside Ethereum’s native transaction format, including UserOperations, bundlers, an EntryPoint, and paymasters. EIP-8141 aims to bring comparable flexibility more directly into the transaction structure itself.
At heart, it addresses a question that should have been answered long ago: why must an ordinary user understand gas before they can send money?
When you transfer money to a friend through a mainstream payment app, it does not tell you to visit another platform and buy a separate “settlement token” first.
Yet that has long been normal in crypto. You need ETH on Ethereum, SOL on Solana, and BNB on BNB Chain. Move to a new L2, and your first question may again be whether your wallet has the right token to get started. An experienced user might solve the problem with a quick bridge transfer. For someone new, it can be enough to make them close the page.
Frame Transactions could also address other persistent frustrations.
Today, swapping a new token on a DEX often requires one transaction to approve the token and another to execute the swap. If the swap fails, the approval may remain active onchain.
With frames, the preceding approval would be rolled back too, rather than being left behind as a lingering security risk.
The broader possibility is that authorization, execution, fee payment, and verification no longer have to be inseparable parts of a single account’s action. They can be arranged to suit the transaction.
3. Is This Good or Bad for ETH?
This debate has produced two opposite predictions. One says, “If users no longer need ETH for gas, ETH loses its value.” The other says, “Lower friction will make transaction volume soar, so this must be hugely positive for ETH.”
It is too early to draw either conclusion from a proposed change to Ethereum’s transaction architecture. What we can say is that EIP-8141 does not remove ETH’s role as the asset used to pay Ethereum’s underlying network fees.
Users might not see ETH in the wallet UI, but the paymaster, app, or other account covering their transaction would still need to meet an ETH-denominated fee obligation. Frames do not replace the EIP-1559 fee market with a stablecoin fee market.
What changes is where the demand for ETH sits.
Today, a million users might each keep a little ETH in their wallets for gas. In a future with widespread sponsorship, a smaller number of paymasters, wallet providers, and apps might instead manage larger ETH balances and handle fees for those users.
That does not mean every stablecoin fee paid by a user will trigger an immediate purchase of the equivalent amount of ETH on an exchange. Sponsors can hold inventory, replenish it over time, hedge their exposure, or use other treasury strategies. “Users can pay fees in stablecoins” does not, by itself, tell us how much new demand for ETH will appear.
The decisive question is whether a better experience leads to more genuine use of Ethereum.
Imagine that 100 people want to make an onchain transaction. Today, 70 give up while trying to buy ETH for gas, calculate fees, or bridge funds, and only 30 complete it.
If a wallet handles that complexity in the background and 80 people complete their transactions instead, Ethereum could see more activity—and more ETH paid as fees, including the portion burned under EIP-1559—even if those users never hold ETH themselves.
Conversely, if the change brings no additional activity and merely replaces “pay your own fee” with “have someone else pay it for you,” it does not create new value on its own.
Final Thoughts
In my view, EIP-8141 weakens the requirement that every user must keep a small amount of ETH on hand before using Ethereum. Its promise is that removing this hurdle could help more people actually use the network.
Ordinary users do not need to change their habits because of a proposal that has not yet been adopted. But for wallets, the direction is clear.
A good wallet should increasingly let users focus on where their assets are going and what they want to do. It should explain the risks and the final cost. Which account pays the network fee, whether the user is charged in ETH or a stablecoin, and whether a gas sponsor is available are details the wallet should handle in the background whenever it can.
That is often what mature infrastructure looks like.