Ethereum’s layer 2 ecosystem has entered a new phase. Rollups now process more daily transactions than the Ethereum mainnet, making the network’s long-term scaling strategy a measurable reality rather than a roadmap promise.
That milestone settled the transit argument, but still opened a tougher economic debate. Activity has shifted upwards, mainnet fee revenues have weakened, and the market is now wondering whether profits from L2 growth will still flow back into Ethereum or stay with rollup operators and applications.
Activity moves up the pile
The shift is now visible for the largest rollups. The data shows that the tracked L2 networks have processed more transactions per day than the Ethereum mainnet since mid-2024, and the trend is only increasing in the first quarter of 2026.
Base has driven much of that expansion. In early 2026, daily transactions crossed two million on several days, with Arbitrum and OP Mainnet each adding hundreds of thousands of additional transactions. Scroll and zkSync Era also contributed meaningful volume. As a result, the combined L2 ecosystem now processes several multiples of the Ethereum mainnet’s daily number of transactions on active days.
Cheaper rollups have changed the math
The March 2024 Dencun upgrade, especially EIP-4844, accelerated that migration. Blob-carrying transactions reduced data availability costs for rollups, making low-cost execution much more practical for commerce, gaming, and consumer applications.
However, that same efficiency came at a cost to Ethereum’s base layer. Mainnet fee income and $ETH Burn fell sharply after Dencun, and the old model that tied Ethereum’s value directly to expensive on-chain activity weakened. Rollups became cheaper and more useful, yet each transaction contributed far less to base tier revenue than comparable activities in 2021.
Value capture remains unresolved
In the meantime, that leaves the economic question open. Arbitrum still has the largest total L2 value in the source material, with activity concentrated in perpetuals, yield products and liquid staking. Nevertheless, ARB does not automatically capture sequencer revenue under the current structure, so the value of tokenholders still depends on governance choices that remain uncertain.
Base presents another model. Coinbase retains sequencer revenue from Base, giving it one of the clearest corporate structures in the rollup market. Furthermore, this has reinforced the view that many L2s can evolve into business ecosystems built around application revenue rather than just neutral layers of scale.
Recent commentary around MegaETH pointed in the same direction, describing a model in which brand and application economics drive the chain.
Ethereum still performs the role of settlement under that activity. It remains the primary collateral in much of the L2 economy, and it still secures the broader stack. It is striking that the increase in scale now looks much stronger than before. Yet the case of value capture remains more uncertain, especially as sequencer centralization, liquidity fragmentation and bridge risk continue to shape the L2 market.
Related: Ethereum Foundation dumps 10,000 $ETH to BitMine