Ethereum Gas Fees at $0.01: Do Layer 2s Still Matter?

Gas fees are under $0.01, daily transactions have jumped from 1.1M to 1.4M, and TPS has climbed from 15 to nearly 18. Ethereum mainnet in 2025 is performing at its best in history, which puts Layer 2 on the spot: rollups were created to solve “expensive fees, slow speeds,” so what reason do they have left to exist now?

This article will examine why L2s still matter beyond low fees, where their real advantages remain, and how the Ethereum stack can evolve so both mainnet and rollups can win.

The Numbers Don't Lie

Over the past 12 months, Ethereum’s performance profile has changed fast. After the Fusaka upgrade in December 2025, average gas fees dropped to around $0.01, a 99% decrease from the $5+ levels of 2024. Gas limit increased to 60 million, blob capacity expanded, and node software was optimized, setting up a very different baseline for what “normal” mainnet usage looks like in 2025.

The results are hard to ignore. Ethereum mainnet now processes nearly 18 TPS, up from 15 TPS, while daily transactions jumped from 1.1 million to 1.4 million in just a few months. Fees used to be the choke point, with swaps reaching $50-100 in 2021, yet now swaps can land under $0.50 and ETH transfers cost a few cents. 

If mainnet already looks this cheap and this responsive, the obvious next step is to ask what Layer 2 still exists to solve and that takes us straight into the economic tension L2s create for Ethereum itself.

Wall Street Slashes Expectations

Standard Chartered put Ethereum’s new reality into a single brutal frame. In March 2025, Geoff Kendrick, Global Head, Digital Assets Research at Standard Chartered, cut ETH’s price target from $10,000 to $4,000, a 60% reduction, then pinned the downgrade on one claim: L2s are bleeding Ethereum dry. Kendrick argued L2s, especially Base, now extract super profits across the Ethereum ecosystem, estimated Base alone removed $50 billion in market cap from Ethereum, then labeled the moment “Ethereum’s Midlife Crisis” while describing Ethereum as commoditizing itself inside the rollup framework it created.

CoinMetrics numbers sharpen the picture after the Dencun upgrade. Base earned approximately $98 million in transaction fee revenue, yet paid only $4.9 million back to Ethereum mainnet, leaving roughly $94 million retained at the rollup layer. The debate stops revolving around scaling and starts revolving around value capture, which leads straight into the next section on how Dencun rewired L2 rent and why critics describe the post upgrade era as silent plunder.

"The Silent Plunder"

A 21Shares report paints a more concerning picture by tracing where revenue goes after Dencun. Before Dencun in March 2024, L2s paid meaningful rent to Ethereum when posting data to L1. After Dencun, posting costs dropped by 90%, which changed the split: rollups keep most profits instead of sharing much with mainnet.

Ethereum L1 revenue in 2025 fell to roughly $286 million, even lower than post Dencun levels from the previous year. Daily ETH burn dropped from thousands of ETH to around 100 ETH per day, which flipped the “ultrasound money” narrative as ETH moved from deflationary to slightly inflationary. Blockworks captured the contradiction with the label “Ethereum’s Paradox”: usage pushing into new highs while fees keep collapsing, transaction count above 50 million per month, unique active addresses at multi year highs, yet fee revenue falling apart.

Qiao Wang: "Recommending ETH L2 Is Irresponsible"

Qiao Wang, co-founder of Alliance DAO, one of the largest Web3 accelerators, pushed the debate past fees and into builder reality. In early 2025, he sold his entire ETH holdings after 10 years, then tied the decision to a pattern he kept seeing across hundreds of portfolio startups: projects launching on Ethereum L2s struggle massively to attract users, while similar products on other chains perform much better.

His argument focused on distribution, liquidity, and user gravity. Fragmented L2 landscapes split attention, split liquidity, and force new teams into uphill battles for discovery, while alternative ecosystems often deliver clearer user funnels and faster feedback loops. He summarized the conclusion in a single line, “Recommending Ethereum L2 is irresponsible.” Alliance DAO data adds weight, showing Ethereum’s share of first time funding applications dropped below 50% in H1 2025, signaling founder allocation shifting before markets fully price the change.

Vitalik: "L2s Are Scaling Ethereum 17x"

Vitalik Buterin acknowledges the tension, then reframes the conversation around architecture, not short term economics. In a January 2025 blog post, he argued L2s in 2025 look far removed from early 2019 experiments. Rollups reached key decentralization milestones, secured billions of dollars in value, and already scaled Ethereum transaction capacity by 17x. He underlined the number directly, 17x, not 17%, positioning rollups as the scaling surface where Ethereum expands while keeping the base layer security model intact.

From Vitalik’s perspective, L2s do not exist as a temporary patch while mainnet catches up. Rollups sit at the center of Ethereum long term design, with a multi network strategy where different L2 technologies serve different use cases rather than one universal chain. He also floated an alignment path: L2s dedicating a portion of revenue to support the wider Ethereum ecosystem through fee burning, staking, or funding public goods, shifting the debate from rollup existence to rollup contribution.

21Shares: "Most L2s Won't Survive Past 2026"

The “State of Crypto” report from 21Shares in December 2025 framed the L2 ecosystem as a market reaching a breaking point after 2 years of rapid expansion. More than 50 L2s are competing, while transaction flow concentrates heavily at the top. Base, Arbitrum, and Optimism account for nearly 90% of total transactions, with Base alone capturing over 60% market share.

Usage trends reinforce a consolidation path. L2 activity is down 61% since June, pushing many smaller networks into zombie chain territory with minimal activity. 21Shares expects a leaner, more resilient set of networks to define Ethereum scaling layer by the end of 2026. Economics lines up with the same outcome: Base was the only profitable L2 in 2025 at around $55 million, while fee wars after Dencun pushed most rollups into losses.

Base: A Success Story and Its Lessons

Coinbase Base works as the cleanest case study for L2 success, and it also sits at the center of the current controversy. Base did not win via marginal tech advantages alone. Base won via distribution, because Coinbase brings 9.3 million monthly active trading users into a single ecosystem, then turns onboarding, liquidity, and discovery into a built in growth loop.

Data from Token Terminal and Bankless:

• 2025 YTD Revenue: $75.4 million (62% of all L2s)
• DeFi TVL: $4.63 billion (46% of L2 market)
• Ecosystem applications: $369.9 million in revenue

Those numbers explain why Base keeps pulling share, but they also expose the tension inside Ethereum scaling economics. Base collected $98 million and paid Ethereum $4.9 million, a 5% ratio, leaving most profit capture at the rollup layer while ETH holders see limited direct benefit. James Beck, Head of Growth at ENS Labs, captured the core complaint in one line: “Ethereum is a neutral verification layer, but Ethereum mainnet is not being fairly compensated for the work it is doing.”

The Real Value of L2s: Not "Cheap Fees"

Low fees no longer carry the rollup narrative, so the value case moves to capability. L2s keep a place by delivering execution environments Ethereum mainnet does not target, then turning speed, throughput, and customization into a product builders can ship and users can feel.

1. Extreme Speed for Gaming

Ethereum mainnet runs near 12s block times, which creates noticeable latency in real time gameplay. Base launched Flashblocks in July 2025 and cut block time from 2s to 200ms. MegaETH pushes toward 100,000 TPS with 10ms blocks. Immutable X processes over 4,000 TPS with near instant latency. These profiles match gaming requirements where responsiveness drives retention.

2. Throughput for High Frequency DeFi

L2s combined now process over 1.9 million transactions per day, with Base leading near 40% across Ethereum ecosystem capacity. High frequency strategies need dense bursts, rapid confirmation loops, and predictable execution windows. Running 1,000 transactions in 1 second still aligns more naturally with rollup design.

3. Enterprise Customization

Enterprise adoption is turning rollups into configurable distribution layers: Sony launched Soneium, Kraken launched INK, Uniswap has UniChain, Robinhood integrated Arbitrum and is developing its own L2. 21Shares predicts exchange backed L2s become 1 of 3 pillars shaping the next phase, alongside ETH aligned designs like Linea and high performance entrants like MegaETH, since brands want control over UX, compliance surfaces, and ecosystem incentives.

4. Institutional Adoption

Institutional adoption remains the long game because global finance prioritizes security, neutrality, and credible settlement. In a Bankless debate in June 2025, Ryan Berckmans framed institutional adoption as the big prize, arguing the global economy wants to come on chain, risk defines what can scale, and Ethereum fits as the right home. Ethereum still holds over 80% market share in tokenized RWAs, reinforcing the pull toward security and neutrality across the Ethereum ecosystem.

Across these cases, L2 value comes from specialization, not discounts. Gaming demands sub second responsiveness. High frequency DeFi demands burst throughput. Enterprises demand configurable execution plus distribution. Institutions demand credible settlement. Cheap L1 fees retire the old pitch and leave a sharper question for the next section: when execution lives on L2 and settlement lives on L1, how can incentives align so both layers win.

The Future: Settlement Layer vs Execution Layer

Ethereum is moving toward a split-stack model where mainnet anchors finality and security while rollups handle high-frequency execution. Blockworks frames value around credible neutrality, security, and settlement across L2s rather than maximizing L1 fee intake, which helps explain the current divergence: fee income compresses while Ethereum still holds $70 billion in TVL, far above Solana at $9.3 billion and BNB Chain at $7.18 billion.

In this table, we can see cheap L1 fees change the sales pitch. Mainnet now concentrates on credible neutrality, security, and settlement finality, while L2s concentrate on fast execution and higher throughput, which leaves one unresolved issue at the center of the stack: economic alignment. 

Naturally, this split pushes one issue to the center of the stack: who pays, and how value returns to Ethereum. Standard Chartered argues alignment needs a hard policy tool, proposing an Ethereum Foundation “tax” on L2 super-profits and warning continued pressure on ETH-BTC without a new value-return path. Vitalik argues for a softer route, pushing voluntary contributions through fee burning, staking, or public goods funding. Fusaka introduces a protocol level lever through EIP-7918, expected to set a blob fee floor so Ethereum captures minimum revenue tied to rollup activity, and Analyst Yi predicts this mechanism could lift ETH burn rate by 8x in 2026.

Conclusion: Do L2s Still Have Value?

Yes, L2s still have value, and the 2025 case looks different from the early scaling era. Mainnet fees near a few cents move the narrative away from “fee relief” and toward purpose built execution environments where speed, throughput, and configurability define product quality.

Gaming demands millisecond latency, high frequency DeFi demands sustained TPS headroom, enterprises demand compliance surfaces plus chain level customization, and onboarding toward billions of users demands rollup style execution at scale. Rollups stay relevant by matching these requirements rather than competing on marginal cost alone.

Selection pressure shapes the landscape next. 21Shares expects consolidation through 2026, leaving a smaller set built around durable advantages such as distribution via Base and Coinbase, ETH aligned economics via Linea and fee burning, and high performance execution via MegaETH and near real time throughput. The urgent question now sits in incentives: How can Ethereum mainnet and L2 networks align value creation with value return, and which economic designs will define winners across the stack in 2026?