Key Takeaways
- Most of what you pay on a Layer-2 network is the cost of publishing transaction data back to Ethereum, not the cost of running your transaction on the L2 itself.
- Data availability upgrades gave rollups a dedicated, cheaper space to post that data, which is the main reason L2 fees fell rather than any single network being clever.
- A fair fee comparison has to account for blob market pricing, batch sizes, and how each network passes savings on to users — headline averages hide a lot.
- Lower fees come with trade-offs around data retention, sequencer centralization, and how quickly funds can exit, so cheaper is not automatically safer.
If you have used an Ethereum Layer-2 network recently and noticed it costs far less than it used to, the reason is mostly structural. A Layer-2 (L2) is a separate chain that processes transactions off the main Ethereum chain and then posts a compressed record back to Ethereum for security. For most of their history, the biggest expense in that process was not running the transactions. It was paying to store the compressed record on Ethereum, where block space is scarce and expensive.
Data availability upgrades changed that equation. Instead of competing for the same expensive space used by every other Ethereum transaction, rollups got a separate, cheaper lane built specifically for the data they need to publish. Understanding that shift is the key to reading L2 fees honestly, and to spotting when a network is genuinely cheap versus when it is just temporarily quiet.
What you are actually paying for on an L2
A Layer-2 fee is not one number. It is two costs bundled together. The first is the execution cost: the work the L2 does to run your transaction on its own chain. This part is genuinely cheap, because the L2 has lots of capacity and is not fighting for Ethereum block space. The second is the data cost: the price of posting your transaction's compressed data back to Ethereum so anyone can verify and reconstruct the L2's state.
For a long time the data cost dominated. When Ethereum's base layer was busy, the price of posting rollup data climbed alongside ordinary gas prices, so L2 users felt every spike on the main chain even though they were not transacting there directly. This is why two L2s with nearly identical technology could show very different fees on the same day. They were both at the mercy of the same data publishing market.
How data availability upgrades cut the bill
Data availability refers to the guarantee that the data behind a transaction is published somewhere anyone can access it, so the network's state can always be checked and rebuilt. Rollups depend on this guarantee; without it, you could not prove the L2 was behaving honestly. The problem was that storing this data permanently in regular Ethereum blocks was overkill and expensive.
The major upgrade introduced a separate type of data storage often called a blob. A blob is a chunk of data attached to an Ethereum block that rollups can use to publish their records, but which the network only keeps for a limited window rather than forever. That window is long enough for anyone to download and verify the data, but short enough that nodes do not have to store it indefinitely. Because blob space is priced in its own market, separate from ordinary transactions, rollups stopped bidding against every other user for the same scarce space.
The effect on fees was large and direct. Once the most expensive component had its own cheaper venue, the all-in cost of an L2 transaction fell sharply for most networks. Importantly, this was a change to the foundation that all rollups sit on, which is why the drop appeared broadly across the ecosystem rather than at one project at a time.
Why the savings are not identical everywhere
Even with the same underlying upgrade, different L2s pass on savings differently. A network that batches many transactions together before posting spreads the data cost across more users, lowering the per-transaction share. A network with lighter traffic may post smaller, less efficient batches, so each user carries more of the fixed cost. The fee a network charges you also reflects choices it makes: some keep a margin, some subsidize fees to attract users, and some adjust pricing dynamically as the blob market moves.
Comparing L2 fees the honest way
A single "average fee" figure is the least useful way to compare networks, because averages get dragged around by quiet periods, promotional subsidies, and the mix of simple versus complex transactions. If you want a fair monthly comparison, the goal is to separate what each network controls from what the market controls.
| Factor to compare | What it tells you | Why averages can mislead |
|---|---|---|
| Cost of a simple transfer | The floor price for everyday use | Complex contract calls cost far more and skew a blended average |
| Cost of a contract interaction | Real cost for swaps, mints, and DeFi | Networks with little DeFi activity look cheaper than they are in practice |
| Data publishing share | How much of the fee is base-layer data cost | This moves with the blob market, not the L2's own efficiency |
| Batch efficiency | How well the network spreads fixed costs | Low-traffic months raise per-user cost even if the tech is identical |
| Fee policy | Whether the network subsidizes or marks up fees | Subsidies can vanish, making a "cheap" month unrepresentative |
When you track these factors month over month, you can tell apart two very different stories. One is a network getting genuinely more efficient through better batching or compression. The other is a network that simply looks cheap because the shared data market was calm or because it was running a temporary subsidy. Only the first is durable.
The trade-offs behind cheaper fees
Lower fees are not free of consequences, and a careful reader should weigh what changes when data moves to a cheaper venue or off Ethereum entirely. Some networks post their data directly to Ethereum, while others use external data availability layers to cut costs further. Those external options are usually cheaper, but they shift where you place your trust.
- Dramatically lower transaction costs make everyday use, small payments, and experimentation practical.
- A dedicated data market means L2 fees are less exposed to spikes in ordinary Ethereum gas prices.
- Broad, foundation-level savings benefit users across many networks rather than locking value into one chain.
- Cheaper data encourages more applications, which in turn improves batch efficiency and pushes costs lower over time.
- Blob-style data is retained only temporarily, so long-term data availability depends on others archiving it.
- Networks that use external data availability layers trade some of Ethereum's security guarantees for lower cost.
- Many L2s still rely on a centralized sequencer that orders transactions, a separate risk from fee levels.
- Low fees can mask slow or restricted withdrawal paths, which matter most exactly when you want to exit quickly.
None of these trade-offs make L2s a bad choice. They simply mean that the cheapest network on a given day is not automatically the best one for a given purpose. If you are moving large amounts, the exit path and data guarantees matter more than saving a fraction on fees. If you are doing high-volume, low-value activity, raw cost may reasonably win.
What to watch going forward
The direction of travel is clear: as Ethereum expands the amount of dedicated data space available and as rollups get better at compressing and batching, the data portion of fees should keep shrinking. The interesting question is no longer whether L2s can be cheap. It is which networks turn temporary, market-driven savings into lasting structural efficiency, and which ones keep their data verifiable without quietly leaning on a single point of trust.
For anyone tracking this, the practical habit is to compare like with like. Look at the same transaction type on the same day across networks, note how much of the fee is data versus execution, and watch how those shares move month to month. That is the difference between reacting to a headline number and actually understanding what you are paying for.
