Five years later, Vitalik overturned his vision for the future of Ethereum.

Author: Rhythm

On February 3, 2026, Vitalik Buterin said something on X.

This statement caused a stir in the Ethereum community no less than his push for a "Rollup-centric" roadmap in 2020. In that post, Vitalik frankly stated: "Layer 2's original vision as a solution to Ethereum's scalability through 'branded sharding' is no longer valid."

In a single sentence, this almost signaled the end of the mainstream Ethereum narrative of the past five years. The Layer 2 camp, once hailed as Ethereum's savior, is facing its biggest legitimacy crisis since its inception. More direct criticism followed, with Vitalik writing scathingly in a post: "If you create an EVM that processes 10,000 transactions per second, but its connection to L1 is achieved through multisignature bridges, then you are not scaling Ethereum."

Why has what was once a lifeline become a burden to be discarded? This is not merely a shift in technological direction, but a brutal struggle over power, interests, and ideals. The story begins five years ago.

How did Layer 2 become Ethereum's lifeline?

The answer is simple: it's not a technological choice, but a survival strategy. Back in 2021, Ethereum was mired in the "aristocratic chain" quagmire.

The data doesn't lie: On May 10, 2021, Ethereum's average transaction fee reached an all-time high of $53.16. At the height of the NFT craze, gas prices soared to over 500 gwei. What does this mean? A typical ERC-20 token transfer might cost tens of dollars, while a token swap on Uniswap could cost as much as $150 or more.

The DeFi Summer of 2020 brought unprecedented prosperity to Ethereum, with the total value locked (TVL) surging from $700 million at the beginning of the year to $15 billion by the end, an increase of over 2100%. However, this prosperity came at the cost of extreme network congestion. In 2021, as the NFT wave swept in, the minting and trading of blue-chip projects like Bored Ape Yacht Club exacerbated the network's problems, with gas fees for a single NFT transaction often reaching hundreds of dollars. One collector in 2021 received offers exceeding 1000 ETH for a Boring Ape figure, but ultimately gave up due to the exorbitant gas fees and complex transaction process.

Meanwhile, a challenger called Solana emerged. Its data was astonishing: tens of thousands of transactions per second with transaction fees as low as $0.00025. The Solana community not only mocked Ethereum's performance but also directly attacked its bloated and inefficient architecture. The notion that "Ethereum is dead" was rampant, and anxiety filled the community.

It was against this backdrop that in October 2020, Vitalik formally proposed a concept in "An Ethereum Roadmap Centered on Rollups": positioning Layer 2 as a "branded shard" of Ethereum. The core of this concept is that Layer 2 processes massive amounts of transactions off-chain, then packages the compressed results back to the mainnet, thereby theoretically achieving infinite scalability while inheriting the security and censorship resistance of the Ethereum mainnet.

At that point, the future of the entire Ethereum ecosystem was almost entirely pinned on the success of Layer 2. From the Dencun upgrade in March 2024, which introduced EIP-4844 (Proto-Danksharding) specifically to provide cheaper data availability space for Layer 2, to various core development meetings, everything was paving the way for Layer 2. After the Dencun upgrade, the cost of data publishing in Layer 2 decreased by at least 90%, and Arbitrum transaction fees plummeted from approximately $0.37 to $0.012. Ethereum attempted to gradually push L1 into the background, positioning it as a quiet "settlement layer."

But why wasn't this bet fulfilled?

Those "centralized databases" valued at $1.2 billion

If Layer 2 had truly achieved its initial vision, it wouldn't be out of favor today. But the question is, what exactly did it do wrong?

In his article, Vitalik incisively points out the fatal flaw: the progress of decentralization is too slow. The vast majority of Layer 2 blockchains have yet to reach Stage 2—possessing a fully decentralized system for proving fraud or validity, and allowing users to withdraw assets without permission in emergencies. They are still controlled by centralized sequencers that manage the packaging and ordering of transactions, essentially resembling centralized databases disguised as blockchains.

The conflict between commercial reality and technological ideals is starkly exposed here. Take Arbitrum as an example: its developer, Offchain Labs, secured $120 million in its Series B funding round in 2021, valuing the company at a staggering $1.2 billion, with investors including top-tier institutions like Lightspeed Venture Partners. Yet, to this day, this behemoth, boasting over $15 billion in locked funds and holding approximately 41% of the Layer 2 market share, remains stuck in Stage 1.

Optimism's story is equally intriguing. This project, led by Paradigm and Andreessen Horowitz (a16z), completed a $150 million Series B funding round in March 2022, bringing its total funding to $268.5 million. In April 2024, a16z even privately purchased $90 million worth of OP tokens. Yet, even with such substantial capital backing, Optimism only reached Stage 1.

The rise of Base reveals another dimension of the problem. As a Layer 2 blockchain launched by Coinbase, Base quickly became a market darling after its mainnet launch in August 2023. By the end of 2025, Base's TVL had reached $4.63 billion, accounting for 46% of the entire Layer 2 market, surpassing Arbitrum to become the Layer 2 with the highest DeFi TVL. However, Base has a lower degree of decentralization because it is entirely controlled by Coinbase, making its technical architecture closer to a centralized sidechain.

Starknet's story is even more ironic. This Layer 2 platform, built on ZK-Rollup technology and developed by Matter Labs, has raised a total of $458 million, including a $200 million Series C funding round in November 2022 led by Blockchain Capital and Dragonfly. However, the price of its token, STRK, has shrunk by 98% from its all-time high, with a market capitalization of approximately $283 million. According to on-chain data, its daily protocol revenue is insufficient to even cover the operating costs of a few servers, and its core nodes remain highly centralized, not reaching Stage 1 until mid-2025.

Some project teams have even privately admitted that they may never be fully decentralized. Vitalik cited an example in a post: a project argued that they would never go further into decentralization because "customers' regulatory needs require them to retain ultimate control." This infuriated Vitalik, who responded bluntly:

"This might be doing the right thing for your customers. But clearly, if you do this, then you're not 'scaling Ethereum'."

This comment is practically a death sentence for all projects that claim to be Ethereum L2 but refuse decentralization. What Ethereum needs is a clone that can extend decentralization and security to a wider space, not a group of appendages that wear the Ethereum cloak but practice centralized operations.

A deeper issue lies in the irreconcilable conflict between decentralization and commercial interests. A centralized ordering mechanism means project teams can control MEV (Maximum Extractable Value) revenue, respond more flexibly to regulatory requirements, and iterate products more quickly. Complete decentralization, on the other hand, means relinquishing this control, handing power over to the community and validator network. For projects with venture capital funding and pressure to grow, this is a difficult choice.

If Layer 2 truly achieves complete decentralization, will it still fall out of favor? The answer is likely still yes. Because Ethereum itself has changed.

When the mainnet is faster and cheaper than sidechains

Why does Ethereum no longer need Layer 2 for scaling?

Back on February 14, 2025, Vitalik released a crucial signal. He published an article titled "Even in L2-heavy Ethereum, there are reasons to have higher L1 gas caps," explicitly stating that "L1 is scaling." At the time, this sounded more like a consolation to mainnet fundamentalists, but in retrospect, it was actually the clarion call for the Ethereum mainnet to begin competing with Layer 2 again.

Over the past year, Ethereum's L1 scaling has far exceeded everyone's expectations. Technological breakthroughs have come from multiple dimensions: EIP-4444 reduced the storage requirements for historical data, stateless client technology made node operation more lightweight, and most importantly, the continuous improvement of the Gas Limit. At the beginning of 2025, Ethereum's Gas Limit was 30 million; by mid-year, it had increased to 36 million, a 20% increase. This is the first significant increase in Ethereum's Gas Limit since 2021.

But this is just the beginning. According to the Ethereum core developers' plans, two major hard fork upgrades will take place in 2026. The Glamsterdam upgrade will introduce perfect parallel processing capabilities, with the Gas Limit soaring from 60 million to 200 million, an increase of more than three times. The Heze-Bogota fork will add the FOCIL (Fork-Choice Enforced Inclusion Lists) mechanism, further improving block building efficiency and censorship resistance.

The Fusaka upgrade, completed on December 3, 2025, has already demonstrated the power of L1 scaling to the market. After the upgrade, Ethereum's daily transaction volume increased by approximately 50%, the number of active addresses rose by approximately 60%, and the 7-day moving average of daily transaction volume reached a record high of 1.87 million, surpassing the record set during the DeFi peak in 2021.

The results are astonishing: Ethereum mainnet transaction fees have plummeted to extremely low levels. In January 2026, the average Ethereum transaction fee dropped to $0.44, a decrease of over 99% from the peak of $53.16 in May 2021. During off-peak hours, the cost of a transaction is often below $0.10, sometimes even as low as $0.01, with gas prices as low as 0.119 gwei. This figure is approaching the level of Solana, and Layer 2's biggest cost advantage is being rapidly eroded.

In his February article, Vitalik did a detailed calculation. He assumed an ETH price of $2,500, a gas price of 15 gwei (long-term average), and a demand elasticity close to 1 (meaning doubling the gas limit would halve the price). Under this assumption:

Censorship resistance requirement: Currently, enforcing a transaction subject to L2 censorship via L1 requires approximately 120,000 gas and costs $4.50. To reduce the cost to below $1, L1 would need to scale up 4.5 times.

Cross-L2 asset transfer: Currently, withdrawing from one L2 to L1 requires approximately 250,000 gas, and depositing into another L2 requires 120,000 gas, for a total cost of $13.87. With an ideal optimized design, it would only require 7,500 gas, costing $0.28. To reach the target of $0.05, a scaling up of 5.5 times is needed.

Large-scale exit scenario: Taking Sony's Soneium as an example, PlayStation has approximately 116 million monthly active users. If an efficient exit protocol is used (7,500 gas per user), Ethereum can currently support the emergency exit of 121 million users within a week. However, to support multiple applications of this scale, L1 needs to scale by approximately nine times.

These expansion goals are being gradually achieved by 2026. Technological advancements have fundamentally changed the game. When L1 can become fast and cheap on its own, why should users tolerate the cumbersome cross-chain bridging, complex user experience, and potential security risks of Layer 2?

The security concerns surrounding cross-chain bridges are not unfounded. In 2022, cross-chain bridges became a major target for hackers. In February, the Wormhole bridge suffered a $325 million theft; in March, the Ronin bridge experienced the largest DeFi attack in history, losing $540 million; and bridging protocols such as Meter and Qubit were also compromised. According to Chainalysis, the total amount of cryptocurrency stolen from cross-chain bridges in 2022 reached $2 billion, accounting for the majority of all DeFi attack losses that year.

Liquidity fragmentation is another pain point. With the surge in the number of Layer 2 blockchains, DeFi protocol liquidity is scattered across more than a dozen different chains, leading to increased slippage, reduced capital efficiency, and a deteriorating user experience. A user wanting to move assets between different Layer 2 blockchains must go through a complex bridging process, wait for lengthy confirmation times, and incur additional fees and risks.

This leads to the next, and most brutal, question: what should those Layer 2 projects that have raised huge sums of money and issued tokens do now?

Valuation Bubble and Ghost Cities

Where did all the money in Layer 2 go?

Over the past few years, the Layer 2 space has resembled a massive financial game rather than a technological revolution. Venture capital firms have been waving checks, pushing the valuations of L2 projects to astonishing heights. zkSync has raised a total of $458 million, Offchain Labs, the company behind Arbitrum, is valued at $1.2 billion, Optimism has raised $268.5 million, and Starknet has raised $458 million. Behind these figures are top-tier venture capital firms such as Paradigm, a16z, Lightspeed, and Blockchain Capital.

Developers are keen on creating complex DeFi Lego sets by nesting different L2 blockchains to attract more liquidity and airdrop hunters. Meanwhile, real users are being worn down by the cumbersome cross-chain operations and high hidden costs.

A harsh reality is that the market is becoming increasingly concentrated at the top. According to data from crypto research firm 21Shares, the three major L2 cryptocurrencies—Base, Arbitrum, and Optimism—control nearly 90% of the trading volume. Base, leveraging Coinbase's traffic advantage and user base, experienced explosive growth in 2025, with its TVL (TVL) soaring from $1 billion at the beginning of the year to $4.63 billion by the end, and quarterly trading volume reaching $59 billion, a 37% increase quarter-over-quarter. Arbitrum maintained its second-place position with a TVL of approximately $19 billion, followed closely by Optimism.

However, outside the top performers, most L2 projects, after losing the incentive of airdrops, saw their real user base plummet, becoming veritable "ghost cities." Starknet is a prime example. Despite its token price falling 98% from its peak, its price-to-earnings ratio remains extremely high relative to its very low daily active users and transaction fee revenue, indicating a significant gap between market expectations for its future and its current ability to create real value.

Ironically, as Layer 2 fees plummeted due to EIP-4844, the data availability fees they paid to L1 also dropped sharply, further reducing Ethereum L1's fee revenue. In January 2026, analysis indicated that the Dencun upgrade caused a large number of transactions to migrate from L1 to the cheaper L2, a major reason why Ethereum network fees fell to their lowest level since 2017. While reducing its own costs, Layer 2 is also draining the economic value of L1.

In its 2026 Layer 2 Outlook report, 21Shares predicts that most Ethereum Layer 2 projects may not survive into 2026, and the market will undergo a brutal consolidation, with only high-performance, truly decentralized projects with unique value propositions ultimately surviving.

This is precisely Vitalik's true intention in launching this attack. He wants to burst the bubble of this infrastructure self-congratulatory frenzy and pour cold water on this unhealthy market. If a Layer 2 cannot provide more interesting and valuable features than Layer 1, then it will ultimately only become an expensive transitional product in the history of Ethereum's development.

Ethereum is regaining its sovereignty

Vitalik's latest suggestion points to a new path for Layer 2: abandoning scaling as the sole selling point and instead exploring functional added value that Layer 1 cannot or is unwilling to provide in the short term. He specifically listed several directions: privacy protection (achieving on-chain privacy transactions through zero-knowledge proof technology), efficiency optimization for specific applications (such as games, social networks, and AI computing), ultra-fast transaction confirmation (milliseconds rather than seconds), and exploration of non-financial use cases.

In other words, Layer 2 will transform from an extension of Ethereum into various plugins. They will no longer be the sole savior for scaling, but rather a functional extension layer within the Ethereum ecosystem. This is a fundamental shift in positioning and a return of power—Ethereum's core values ​​and sovereignty will be re-anchored to L1.

Vitalik also proposed a new framework: viewing Layer 2 as a spectrum rather than a binary classification. Different L2s can have different trade-offs in terms of decentralization, security guarantees, and functionalities. The key is to clearly explain to users what guarantees they provide, rather than all claiming to be "extending Ethereum."

This reckoning has begun. Those Layer 2 projects sustained by exorbitant valuations but lacking any real daily active users are facing their final judgment. Those projects that can find their unique value proposition and truly achieve decentralization may survive in the new landscape. Base may continue to maintain its lead by leveraging Coinbase's traffic advantage and its ability to attract Web2 users, but it needs to address criticisms regarding its insufficient decentralization. Arbitrum and Optimism need to accelerate their Stage 2 development to prove they are more than just centralized databases. ZK-Rollup projects like zkSync and Starknet need to demonstrate the unique value of their zero-knowledge proof technology while significantly improving user experience and ecosystem prosperity.

Layer 2 hasn't disappeared, but its era as Ethereum's only hope is over. Five years ago, cornered by competitors like Solana, Ethereum pinned its hopes for scaling on Layer 2 and restructured its entire technology roadmap for this purpose. Five years later, it has discovered that the best scaling solution is to make itself more powerful.

This isn't betrayal, but growth. Those Layer 2 projects unable to adapt to this evolution will pay the price. When the Gas Limit surges to 200 million by the end of 2026, when Ethereum L1 transaction fees stabilize at a few cents or even lower, and when users find they no longer need to endure the complexity and risks of cross-chain bridges, the market will vote with its feet. Those projects that once commanded exorbitant valuations but failed to create real value for users will be forgotten in this great sifting process.