If you've used any Layer 2 rollups in the past year, such as Arbitrum, Optimism, Base or Starknet, you've likely seen a dramatic drop in your transaction fees. Where a swap would typically cost a few dollars on L2, it now can cost as little as one cent. This drop in costs is not accidental. EIP-4844, which went into effect on March 13, 2024, as part of the Dencun hard fork of Ethereum, is what caused it. The update to Ethereum is known as Proto-Danksharding and it has introduced a brand new transaction type, called "blob transaction", that has fundamentally changed how Layer 2 networks send data back to Ethereum. As a result, L2 transaction fees have decreased by between 10x and 100x, depending on the rollup used. The infrastructure change gives you a better experience as an owner or user of a cryptocurrency when using any Ethereum L2. <h2>The Problem: Why L2 Fees Were Still Expensive</h2> Layer 2 rollups were always meant to be Ethereum's way to scale. They do off-chain transactions, bundle them into a group (or batch) of transactions, send this batch of transactions back to Ethereum's mainnet for validation, and send a compressed version of that batch back to L1 through the Ethereum blockchain. The last piece of this process was the issue. Before EIP-4844 came into play, rollups needed to submit their batches of data as "calldata." Calldata is an on-chain data type that is stored permanently on Ethereum's execution layer. In addition to being stored permanently, every byte of call data costs gas, so the more calldata is used by rollup operators, the greater the number of transactions happening on standard Ethereum. Rollups therefore needed to compete with standard Ethereum transactions for the same block space, and as a result, their transaction fees were increasing along with the demand for transactions on Ethereum. In addition, the vast majority of rollups' operational costs were incurred during the posting of data to Ethereum's L1. The cost floor for these costs was the amount that rollups paid to Ethereum to post their batch data permanently, regardless of how low L2 transaction fees were compared to mainnet transaction fees. Ultimately, both Ethereum's developers and the broader cryptocurrency community realized that rollup data should not be stored on-chain continuously; rather, it should only be there to help verify that L2 transactions occurred.<a href="https://eips.ethereum.org/EIPS/eip-4844" rel="noopener noreferrer" target="_blank"> </a><a href="https://eips.ethereum.org/EIPS/eip-4844" rel="noopener noreferrer" target="_blank">EIP-4844</a> reflects the acknowledgment of this fact. <h2>What Are Blob Transactions?</h2> EIP-4844 introduced a new transaction type called Type 3 which adds a new type of data called a blob. A binary large object (blob) is a fixed-size chunk of data about 128KB In size. Each blob is made up of 4096 field elements, or "words," where each word is 32 bytes in length. Blob data will be broadcast to the consensus layer (the Beacon Chain) and are stored for a period of<a href="https://consensys.io/blog/ethereum-evolved-dencun-upgrade-part-5-eip-4844" rel="noopener noreferrer" target="_blank"> </a><a href="https://consensys.io/blog/ethereum-evolved-dencun-upgrade-part-5-eip-4844" rel="noopener noreferrer" target="_blank">approximately 18 days</a> on beacon nodes, after which they will be automatically deleted, as opposed to being stored on the execution layer of Ethereum permanently. This is a significant difference in design; blobs provide the same data availability guarantee that rollups need without the cost associated with permanent storage. The verification window allows for sufficient time for both provers and verifiers to analyze the data and provide proof of authenticity or fraud. By the end of the verification window the data will be purged, significantly reducing the overall storage burden of the network, allowing blob transactions to be charged significantly less than call data. <h2>The Separate Fee Market: Blobs Don't Compete for Block Space</h2> EIP-4844 created a separate fee market for blob transactions that has nothing to do with the standard Ethereum network's gas fees. Since blob transactions will never compete for block space; blobs won't ever compete for the same block as DeFi swaps, NFT mints, or any other mainnet traffic. The fact that blob fees are paid in a different measuring unit (blob gas) ensures that blob fees will remain distinct from the effects of meme coins that can raise the Ethereum L1 gas price to 200 gwei. There will be up to 6 blobs in a block (3 is the target). This approach to setting a target fee for blobs is equivalent to the EIP-1559 method; as the number of blobs consumed exceeds the target, then the blob fee will rise, and it will drop if below target consumption. As a result, the potential future price of rollups will be predictable and self-regulating, and the operation of L2 will generally remain independent of fluctuations in gas prices on the mainnet. This means that L2 end-user transaction costs will continually decrease. <h2>The Real-World Impact on L2 Fees</h2> After Dencun, the statistics tell the whole story: Base, Coinbase's Layer 2 system, received a<a href="https://www.theblock.co/post/283056/base-dencun-transactions-users-surge" rel="noopener noreferrer" target="_blank"> </a><a href="https://www.theblock.co/post/283056/base-dencun-transactions-users-surge" rel="noopener noreferrer" target="_blank">224% increase in transaction volume</a> after its upgrade and all due to the transactions being cheaper, which attracted a large influx of new customers as a result. Transaction fees on virtually every major rollup have decreased from dollar amounts down to fractions of a penny. Before EIP-4844, an Optimism transaction cost anywhere from $0.50 to $2.00 depending on L1 congestion and frequently<a href="https://thedefiant.io/news/cefi/transaction-fees-on-layer-2-plummet-post-dencun" rel="noopener noreferrer" target="_blank"> </a><a href="https://thedefiant.io/news/cefi/transaction-fees-on-layer-2-plummet-post-dencun" rel="noopener noreferrer" target="_blank">dropped to below $0.01</a> following the upgrade. This is a structural change that allows L2s to compete with lower-cost alternative Layer 1 blockchains such as Solana, all while maintaining the security guarantees associated with Ethereum.The fee reduction has also changed the way that on-chain activity occurs. DeFi protocols, NFT marketplaces, and gaming applications that previously couldn't be supported by L2s due to the high costs of conducting many microtransactions can now be supported. The cost of trading being less than a penny opens up a tremendous amount of design space for building crypto applications. <h2>The Cryptography Behind It: KZG Commitments</h2> Blobs are securely added to the Beacon Chain not just as raw files through a KZG (Kate-Zaverucha-Goldberg) polynomial commitment, which is a validated method for verifying data availability without needing to download the entire blob itself. Utilising KZG's method, a blob's data is cryptographically validated (i.e., proven to be correct) as well as made available (i.e., can be used) via a concisely generated proof via the polynomial. The ability to use data availability sampling is extremely valuable because it allows the nodes to examine only small random samples of blob's data (versus detailing all the details) in order to determine whether or not the blob's data has been made available. Although this cryptographically validated structure is beneficial towards additional Proto-Danksharding solutions based off of Ethereum's eventual plans for<a href="https://ethereum.org/roadmap/danksharding/" rel="noopener noreferrer" target="_blank"> </a><a href="https://ethereum.org/roadmap/danksharding/" rel="noopener noreferrer" target="_blank">fully Danksharding</a>, it will also continue to aid other Ethereum enhancements towards developing a solution allowing hundreds of rollups to be processed concurrent to one another at up to 100,000 tps by subsequently increasing inclusion of blobs per block from the current architecture's six per block to ultimately 64 or more. EIP-4844 has established transaction structure, transaction validation, and transaction fee market (e.g., transaction costs) so that the future added enhancements increase capacity but do not require changes to the system architecting. <h2>What Comes Next: From Proto to Full Danksharding</h2> The endgame was never EIP-4844. The "Proto" refers to the prototype, and a number of follow-up ideas are already being developed. EIP-7691 recommends increasing the number of blobs per block from three to four. In order to encourage the remaining rollups to embrace blobs, EIP-7623 suggests raising the cost of calldata gas. Building on this framework, the<a href="https://blockworks.co/news/ethereum-dencun-upgrade-mainnet" rel="noopener noreferrer" target="_blank"> </a><a href="https://blockworks.co/news/ethereum-dencun-upgrade-mainnet" rel="noopener noreferrer" target="_blank">Pectra update</a> progressively increases Ethereum's data throughput while maintaining controllable node requirements for decentralization. The path is obvious: Blobs are the data layer that enables Ethereum's rollup-centric scaling plan. Whether they are aware of it or not, every crypto user that interacts with an L2 nowadays is already taking advantage of EIP-4844. The decision to cease keeping rollup data permanently and start considering it as what it always was—temporary evidence that the math worked out—is what led to the sub-cent transactions, the quicker settlement, and the<a href="https://coinmetrics.substack.com/p/state-of-the-network-issue-262" rel="noopener noreferrer" target="_blank"> </a><a href="https://coinmetrics.substack.com/p/state-of-the-network-issue-262" rel="noopener noreferrer" target="_blank">surge in on-chain activity</a>.<h1> </h1>

EIP-4844 introduced blob transactions that let Ethereum L2s post data cheaply without competing for block space, cutting rollup fees 10–100x and enabling sub-cent transactions.

EIP-4844 and Proto-Danksharding – Explaining How Ethereum's "Blob" Transactions Reduced L2 Fees

/media/contribution/care730a226381a45a098827a8d57c9e2fd/8lb_0330024447087.stxt

Africa Crypto Taxation in 2026: What Traders in Emerging Markets Are Getting Wrong

/media/contribution/carb8b0146954c24b7a872d53bdf30d8eee/ak7_0330024815180.stxt

Investing Is Not About Who Earns More, but Who Lasts Longer: Essential Position Management Survival Rules for Beginners

/media/contribution/car71b7af50446c4d2abc0492ee8bb77ecb/cor_0330023232905.stxt

The ETH OG Who Went Silent Is Buying Again at $19.5 Million Worth

/media/contribution/car31bf3936705049cd8d2ce5828e38da05/363_0330023735807.stxt

Larry Fink’s 2026 Chairman’s Letter: Tokenization and the Fight Against Financial Inequality

/media/contribution/car5abe02d35a114fed9a810f7046df961d/5xf_0330031353546.stxt

Smart Money Is Buying Bitcoin Again: What You Should Know

/media/contribution/car45b2e61875c84ff3848434ee5616d3a9/ufd_0330030732421.stxt