Ethereum co-founder Vitalik Buterin has highlighted Bitcoin’s Utreexo research as an inspiration for Ethereum’s proposed scaling direction, with a potential future architecture combining UTXO-style state, Ethereum’s existing dynamic state and hybrid models.
Buterin discussed the approach in an Aug. 16 post on X, describing it as the “current proposed Ethereum scaling strategy.” His comments indicate that the architecture remains under research and should not be interpreted as a decision to replace Ethereum’s account-based model with Bitcoin’s UTXO system.
The broader objective is to allow Ethereum to handle substantially more activity without dramatically increasing the hardware and storage requirements for running a full node.
Bitcoin’s Utreexo provides a model for reducing state requirements
Utreexo is a Bitcoin research project designed to reduce the amount of state that validating nodes need to store locally.
The concept was introduced by MIT Digital Currency Initiative researcher Thaddeus Dryja in 2019. Instead of requiring every node to maintain the complete Bitcoin UTXO set, Utreexo represents the set using a compact cryptographic accumulator.
Transactions can then include proofs showing that specific unspent outputs exist within the accumulator. Nodes can verify those proofs without maintaining the entire UTXO set locally.
The original Utreexo research paper describes an accumulator whose size grows logarithmically relative to the underlying set.
That idea is relevant to Ethereum because the network faces a similar long-term challenge: more users and applications can mean more persistent state, which can increase the cost of operating and verifying nodes.
Utreexo is not an Ethereum proposal, however. It remains a Bitcoin-oriented research project that illustrates one possible way to separate state availability from the amount of data every node must permanently maintain.
Ethereum researchers explore native UTXO-style payments
Ethereum researchers have been examining whether certain types of activity could benefit from a UTXO-style design without abandoning Ethereum’s existing account model.
A July proposal by Ethereum researcher Toni Wahrstätter explores native UTXOs on Ethereum, particularly for simple payments that do not require persistent smart-contract state.
Ethereum Research: Native UTXOs on Ethereum
The proposal argues that one-time payment workloads could potentially use substantially less permanent state than equivalent account or storage-based transactions.
Under the design, Ethereum would not need to permanently retain the complete payment object in active state. Instead, cryptographic proofs could establish that an output existed, while the network would primarily need to track whether it had already been spent.
The proposal estimates that a system containing one billion such entries could require roughly 300 MB of permanent state, compared with an estimated 100 GB to 150 GB for equivalent account or storage entries.
These figures are design estimates rather than measurements from a live Ethereum implementation.
A hybrid state model could be the bigger idea
Buterin's comments point toward a potentially broader architecture rather than a simple migration from accounts to UTXOs.
Ethereum could theoretically maintain several types of state simultaneously:
Traditional account-based state for smart contracts and applications requiring persistent storage.
UTXO-style state for simple payments and other short-lived transactions.
Hybrid models for applications that need characteristics of both approaches.
Cryptographic accumulators and proofs to reduce how much information nodes need to store locally.
Such a system could allow different workloads to use the state structure that fits them best.
The important distinction is that Ethereum would not necessarily need to choose between its current account model and UTXOs. Instead, researchers are exploring whether the two approaches could coexist.
Recursive STARKs target another Ethereum bottleneck
Buterin has also explored recursive STARKs as a way to reduce the bandwidth required to propagate information through Ethereum's network.
In a January Ethereum Research proposal, he described a recursive-STARK-based mempool in which nodes could periodically combine validity proofs rather than repeatedly transmitting individual large proofs.
Ethereum Research: Recursive STARK-Based Bandwidth-Efficient Mempool
The proposal uses highly optimized STARK proofs of around 128 KB as an assumption and explores a system where proof aggregation could keep additional bandwidth requirements relatively stable as the number of propagated objects increases.
This is different from the native UTXO proposal.
UTXO-style state primarily addresses state-storage and state-management costs, while recursive proofs target verification and network-bandwidth efficiency.
Combining the concepts could theoretically create a more efficient Ethereum architecture, but no such complete system has been approved or scheduled for deployment.
Ethereum is not moving to Bitcoin’s UTXO model
Buterin's comments have generated comparisons between Ethereum and Bitcoin, but the two networks continue to use fundamentally different accounting models.
Bitcoin primarily tracks unspent transaction outputs, while Ethereum uses an account-based state model in which balances, contract storage and other information are maintained as part of the global state.
The research discussed by Buterin does not indicate that Ethereum plans to abandon accounts.
Instead, the emerging direction appears to be selective use of different state models depending on the type of activity.
That could be particularly useful for simple payments, where maintaining persistent account state may be more expensive than necessary.
Hegotá could provide part of the framework
Some of these ideas are also connected to Ethereum's upcoming Hegotá upgrade, although native UTXOs are not currently a confirmed Hegotá feature.
The official Hegotá Meta EIP currently lists FOCIL, or EIP-7805, as the only proposal scheduled for inclusion.
Frame Transactions, or EIP-8141, remains under consideration.
EIP-8141: Frame Transactions
The native UTXO proposal assumes Frame Transactions for its preferred spending design, potentially allowing more programmable approaches to transaction validation, execution and gas payment.
However, this does not mean native UTXOs have been accepted for Hegotá.
Ethereum's official roadmap currently places Hegotá in 2027, following Glamsterdam, which is targeted for Q4 2026.
What this means for Ethereum scaling
The research highlighted by Buterin points toward an Ethereum architecture that could become increasingly modular at the state level.
Instead of requiring every type of transaction to interact with the same persistent state model, Ethereum could eventually use specialized mechanisms for different workloads while relying on cryptographic proofs to keep verification efficient.
If successful, such an approach could help Ethereum increase activity without forcing node operators to continuously increase storage and hardware requirements.
But the technology remains experimental.
There is currently no confirmed Ethereum decision to adopt Utreexo, replace accounts with UTXOs, or deploy native UTXOs as part of Hegotá. Buterin's comments instead highlight a research direction in which Ethereum could borrow ideas from Bitcoin while developing a hybrid model suited to smart contracts and programmable applications.
