Transaction fees on leading Ethereum Layer 2s have dropped so low that the economic model underpinning most of these networks is starting to look genuinely fragile. Base regularly processes swaps for under $0.01. Arbitrum One has seen average fees hover near the same floor for months. Optimism isn’t far behind. The race to zero, which analysts once treated as a distant theoretical outcome, has effectively arrived.
The driver is EIP-4844, which introduced blob-carrying transactions in early 2024. Before blobs, L2s paid Ethereum mainnet fees to post calldata - a cost that put a hard floor under what they could charge users. Blobs created a separate, cheaper data market, and the effect was immediate. L2 fees collapsed within weeks of Dencun shipping, and they’ve stayed low since.
What’s changed more recently is the competitive dynamic between chains. Base, backed by Coinbase, has been aggressive about user acquisition and doesn’t face pressure to generate protocol revenue in the short term. That gives it the luxury of operating near cost - or below it. Arbitrum and Optimism don’t have the same institutional backstop, and their foundations are burning through grant budgets trying to retain developer mindshare against a chain that can effectively subsidise its own growth.
The Sequencer Revenue Problem
L2s currently monetise through sequencer fees - the spread between what users pay and what the network spends posting data to Ethereum. As user fees compress, that spread narrows. Some estimates put Arbitrum’s net sequencer margin below 30% of what it was pre-Dencun, though exact figures vary and the chains don’t publish detailed revenue breakdowns publicly.
Decentralising sequencers, which most L2s have promised but not delivered, makes this harder, not easier. Shared sequencer revenue across a distributed set of operators means each participant earns less, which either requires higher fees or a working token model that compensates validators. Neither is simple.
This Is Still Ethereum’s Win
For Ethereum itself, the L2 fee war is largely a positive story. More activity is settling to mainnet through blob data than ever before, and ETH continues to accrue value as the settlement and staking layer. The pressure falls on L2s to find sustainable business models - whether through app-specific revenue, native token emissions, or eventually MEV capture.
The chains that survive this compression will be the ones that built genuine user lock-in rather than competing purely on cost. Fee minimalism was always a temporary moat.