Home / Wiki / E / EIP-4488

EIP-4488

Cryptocurrency and Digital AssetsUpdated 05 Nov 2026

EIP-4488 is a proposed change to the Ethereum blockchain that would have cut the gas cost of transaction calldata from 16 gas per byte to 3 gas per byte, while capping how much calldata a single block could hold. Vitalik Buterin and Ansgar Dietrichs published it on 23 November 2021 as a quick way to lower fees for rollups, which post their transaction data to Ethereum as calldata.

The proposal was never activated. Its status on the official EIP site is "Stagnant", and the problem it targeted was instead addressed by EIP-4844, known as proto-danksharding, which went live in the Dencun upgrade in March 2024. EIP-4488 is still cited as an example of how Ethereum's developers weigh short-term fee relief against network safety.

01

EIP-4488 in plain terms

Imagine a courier company that charges by the page for every document it stores in a public archive. Smaller courier firms, the rollups, bundle thousands of customers' letters into one thick envelope and pay the archive's page fee. When the page fee is high, every customer's letter costs more.

EIP-4488 proposed cutting the page fee by about four-fifths, but adding a rule that no single delivery truck could carry more than a fixed number of pages. Cheaper pages would help the courier firms, and the cap would stop anyone from overloading a truck and jamming the road.

02

Background: why calldata became expensive

Every Ethereum transaction can carry a field of raw bytes called calldata, which a smart contract reads when it runs. Calldata is stored in the chain's history forever, so the protocol charges for it. Under EIP-2028 (2019) the price is 16 gas for each non-zero byte and 4 gas for each zero byte.

Layer 2 networks such as Arbitrum, Optimism and ZKsync execute transactions off the main chain and post compressed batches back to mainnet so anyone can rebuild their state. Both optimistic rollups and ZK-rollups used calldata for this in 2021, and it made up most of the fee a rollup user paid. During the busy markets of late 2021, high gas fees on layer 1 pushed rollup fees up too.

03

What the proposal changed

EIP-4488 made two linked changes:

  • A lower price. Calldata would cost a flat 3 gas per byte, whether the byte is zero or not.
  • A block cap. A block's total calldata could not exceed 1,048,576 bytes plus a stipend of 300 bytes for each transaction it contains.

The cap mattered because cheaper data, on its own, would let a block filled entirely with calldata grow several times larger than normal. That would strain bandwidth and storage for nodes and could weaken the network's security. The cap kept the worst-case block size close to one megabyte.

The calldata charge for a batch under the existing rules is:

Gascalldata=16×Nnon-zero+4×Nzero\text{Gas}_{\text{calldata}} = 16 \times N_{\text{non-zero}} + 4 \times N_{\text{zero}}

Here Nnon-zeroN_{\text{non-zero}} is the number of non-zero bytes and NzeroN_{\text{zero}} the number of zero bytes. Under EIP-4488 it would simply have been 3 gas times the total byte count.

Worked example: a rollup posts a 100,000-byte batch of which 90,000 bytes are non-zero and 10,000 are zero. Today's rule gives 16 × 90,000 + 4 × 10,000 = 1,480,000 gas. EIP-4488 would charge 3 × 100,000 = 300,000 gas, about 80% less.

04

Why it was not adopted

The proposal changed how blocks are validated, so it needed a hard fork, and client teams were busy with the move to proof of stake (the Merge, September 2022). Some developers also disliked adding a second resource limit to block building and worried that permanent calldata would keep growing the history every node stores. Attention shifted to a cleaner design that gives rollups a separate, cheaper data space.

05

What replaced it: EIP-4844 and later changes

EIP-4844 introduced "blob-carrying transactions". Blobs are large data packets with their own fee market, priced separately from ordinary gas, and nodes delete them after roughly 18 days instead of keeping them forever. Rollups only need the data to be available long enough for anyone to check it. Blobs went live with the Dencun upgrade on 13 March 2024, and rollup fees fell sharply in the following weeks.

The Pectra upgrade in May 2025 went further in the other direction for calldata: EIP-7623 raised its cost for data-heavy transactions, steering bulk data into blobs, while EIP-7691 increased the number of blobs per block. The long-term goal, full danksharding, is part of Ethereum's wider scaling roadmap. For holders of ether (ETH) and users of rollups, the practical outcome is the one EIP-4488 aimed at: cheaper transaction fees on layer 2, reached by a different route.

Last updated 05 Nov 2026