Ethereum L1 and Coinbase's Base L2 abandoned months of joint work on a unified account abstraction standard on September 14, 2026. The two networks will now ship competing protocol-level wallet designs: EIP-8141 (Frame Transactions) for Ethereum's 2027 Hegotá upgrade and EIP-8130 (Keystore-based ...
"While we identified a number of technical solutions, they all required one side or the other to compromise at least a little bit on their core goals. So separate ways we went, putting the burden on wallets to deal with the fragmentation that ensues." — Derek Chiang, Co-author of EIP-8141 and Researcher at Ethlabs
Ethereum L1 and Coinbase's Base L2 abandoned months of joint work on a unified account abstraction standard on September 14, 2026. The two networks will now ship competing protocol-level wallet designs: EIP-8141 (Frame Transactions) for Ethereum's 2027 Hegotá upgrade and EIP-8130 (Keystore-based AA) for Base's Cobalt upgrade, targeted for late September 2026. The split marks the first time Ethereum's largest L2 by user count has formally diverged from L1 on a core transaction format.
The practical fallout lands on wallet developers and application builders, who may need to support two native transaction types across the Ethereum ecosystem. ERC-4337, the existing smart-account standard with over 40 million deployed accounts and more than 170 million UserOperations processed since March 2023, remains as a compatibility bridge — but does not eliminate the integration cost. The economic question is whether two purpose-built standards deliver more aggregate value than one compromised universal standard would have.
On September 14, 2026, Derek Chiang — co-author of EIP-8141 and founding member of Ethlabs (and founder of wallet infrastructure firm ZeroDev) — announced that Ethereum and Base had ended efforts to merge their native account abstraction proposals into a single specification.
The two proposals share the same goal: make account abstraction native to the protocol layer, eliminating the need for the external bundler infrastructure that ERC-4337 currently requires. Both enable gasless transactions, passkey-based authentication, transaction batching, and key rotation without seed phrases. The disagreement was architectural, not functional.
Ethereum will proceed with EIP-8141 (Frame Transactions), co-authored by Vitalik Buterin and nine other researchers and engineers, as a headliner for the Hegotá hard fork. Base will proceed with EIP-8130, authored by Coinbase engineer Chris Hunter, targeting deployment via the Cobalt upgrade.
EIP-8130 introduces a new transaction type paired with an on-chain Keystore. Each transaction names the authenticator contract that validates it, allowing nodes to determine validation requirements without executing EVM code. The design is constrained by intent: nodes filter transactions by looking up the declared verifier rather than simulating wallet logic.
Core capabilities include custom authentication, call batching, gas sponsorship (including fee payment in non-ETH tokens), session keys, sub-accounts, and cross-chain account portability. The proposal was created on October 14, 2025, and is currently in draft status.
EIP-8130 puts flexibility in declared authenticators, Keystore scopes, and execution-time policy managers. Chains can select adoption profiles, meaning not every EVM chain must implement the same subset of features.
EIP-8141 structures transactions as sequences of up to 64 programmable sub-units called "frames." Each frame can perform distinct operations — validation, gas approval, execution — within a single atomic transaction. This design enables full EVM-level programmability during validation.
The proposal was first put forward on January 29, 2026. Co-authors include Buterin, lightclient, Felix Lange, Yoav Weiss, Alex Forshtat, Dror Tirosh, Shahaf Nacson, Derek Chiang, Toni Wahrstatter, and Stavros Vlachakis.
Frame Transactions targets features beyond basic account abstraction: transaction expiry, privacy proofs, signature aggregation, and a pathway to post-quantum authentication. Buterin has argued that separating transaction actions and dependencies could enable parallel validation and reduce costs.
EIP-8130 uses a constrained verifier interface. EIP-8141 uses full EVM execution during validation. As Pedro Gomes, founder of WalletConnect, described it: "EIP-8130 standardizes the product surface directly, while EIP-8141 standardizes a general machine from which developers can construct it."
The breakdown was not over features — both proposals support gasless transactions, passkey wallets, and batching. The breakdown was over values embedded in the architecture.
Ethereum L1 priorities: Censorship resistance, privacy, post-quantum security, and long-term extensibility for the base settlement layer. The Ethereum Foundation's Protocol cluster rated EIP-8141 in its top "S" tier after evaluating 62 candidate EIPs with input from approximately 60 researchers and engineers.
Base priorities: Scalability, configurability, and compliance features suited to commercial applications and enterprise partners. Base processes high transaction volumes and requires enforceable rules that clearly separate permitted transactions from restricted ones.
As Chiang wrote: "Ethereum wanted to be the best version of Ethereum, and Base wanted to be the best version of Base."
The teams explored multiple technical compromises. Each required one side to weaken a core design goal. Neither side accepted the trade-off.
According to data published by Base, EIP-8130 reduces gas consumption for a standard USDC transfer by 63.2% compared to ERC-4337 — from 125,000 gas to approximately 46,000 gas. Transaction byte size drops by 83.4%. Base reports an overall per-transaction cost reduction exceeding 2x compared to its previous smart account implementation.
These figures come from Base's Vibenet development network, where EIP-8130 is currently running. Production performance may vary.
ERC-4337, the existing non-native smart account standard, has processed over 170 million UserOperations since launching on Ethereum mainnet on March 1, 2023. Over 40 million smart accounts have been deployed across Ethereum and Layer 2 networks. By April 2026, total cumulative UserOperations across Ethereum and L2s reached approximately 2.4 billion, according to BundleBear's aggregator.
The standard enabled smart account adoption without requiring any protocol-level changes but carries overhead from its external bundler architecture and the singleton EntryPoint contract.
Account abstraction changes how transaction fees flow through a blockchain network. Under the traditional EOA model, every user must hold ETH to pay gas. Under account abstraction, a sponsor (an application, a paymaster, or a third party) can cover gas on behalf of the user. The user may pay in stablecoins, in the application's token, or not at all.
This matters for fee revenue distribution. When EIP-8141 enables stablecoin gas payments on Ethereum L1, the protocol itself still receives ETH at the base layer — a sponsor wallet fronts the ETH and takes stablecoin compensation from the sender. The economic flow changes at the application layer while preserving the protocol's fee structure.
On L2s like Base, where transaction fees are already low, the question shifts to volume. If native AA reduces per-transaction costs by 63%, the economic rationale depends on whether the cost reduction drives proportionally higher transaction counts. Base's bet is that removing friction — no seed phrases, no ETH requirement, one-tap transactions — converts enough new users to offset per-unit fee compression.
The approximately 200 million existing EOA accounts on Ethereum represent a large installed base that EIP-7702 (shipped in the Pectra upgrade, May 2025) has begun to upgrade. EIP-7702 allows existing EOAs to temporarily delegate to smart contract logic, bridging old and new account types without requiring users to migrate.
The split creates a concrete engineering cost. If both proposals reach production, wallet developers must support two native transaction formats across the Ethereum ecosystem:
Major wallet providers are already adapting. Coinbase Smart Wallet is built around EIP-8130. WalletConnect has partnered with Base and Optimism to implement EIP-8130 across EVM chains. Safe (formerly Gnosis Safe) supports modules for both ERC-4337 and EIP-7702, and would likely extend support to both native standards.
Chiang's assessment is that software can hide the difference from end users. Whether that claim holds depends on execution quality across dozens of independent wallet teams.
| Standard | Network | Vehicle | Status | Target | |----------|---------|---------|--------|--------| | EIP-8130 | Base | Cobalt upgrade | Running on Vibenet devnet | Testnet/mainnet targeted Sept 2026 | | EIP-8130 | OP Stack chains | Post-Cobalt rollout | Planned | Late 2026 | | EIP-8141 | Ethereum L1 | Hegotá hard fork | Scheduled for Inclusion (Aug 27 ACD call) | 2027 | | Glamsterdam | Ethereum L1 | Pre-Hegotá upgrade | Devnet-11 (Sept 14, 84K validators) | Q4 2026 mainnet |
Note: Base's Cobalt deployment timeline carries a caveat. As of mid-September, EIP-8130 remains experimental and runs only on the Vibenet devnet. The September mainnet target may shift.
Ethereum's Hegotá is contingent on Glamsterdam shipping first. Glamsterdam's Sepolia public testnet is tentatively targeted for October 6, 2026, with a Q4 2026 mainnet goal. A further delay into 2027 has been acknowledged as possible.
Pedro Gomes, founder of WalletConnect, publicly sided with EIP-8130 after working on EIP-8141 for months. "After spending months on EIP-8141, I'm convinced EIP-8130 is the better path for native account abstraction. It's simpler, more portable, and focused on what wallets actually need," Gomes wrote on X. He has warned that EIP-8141 in its current form could lead to setbacks for Ethereum due to over-engineering.
Nethermind and Besu client teams raised concerns about EIP-8141's implementation complexity and its impact on Ethereum's validator node complexity budget during the ACD process.
The Ethereum Foundation's position, reflected in its S-tier ranking of EIP-8141, treats the proposal as essential for Ethereum's long-term security posture, particularly its post-quantum authentication roadmap. The Foundation's stated deadline for complete post-quantum security across all layers is December 2029.
The Ethereum-Base account abstraction split is not a technical failure — it is a governance outcome. Two networks with different user bases, different regulatory exposures, and different scaling requirements concluded that a single standard would serve neither well.
The economic implications depend on execution. If Base ships EIP-8130 successfully and the 63% gas reduction drives measurable adoption growth, other OP Stack chains and EVM networks face pressure to adopt it. If Ethereum delivers EIP-8141 with viable post-quantum signatures in Hegotá, it strengthens the case for L1 as the settlement layer of last resort.
The risk is fragmentation without resolution. If wallet teams cannot abstract away the differences, users encounter friction when moving between L1 and L2 — exactly the problem account abstraction was meant to eliminate. Chiang's cautious optimism — "we may well end up with the best possible UX for the end users" — remains conditional on coordinated execution across an ecosystem that just demonstrated it cannot always coordinate.
The approximately $4 trillion crypto market now depends on dozens of independent wallet teams to bridge the gap that ten co-authors and months of negotiation could not close.