What Is EIP-4844 (Proto-Danksharding)? Guide (2026)
— By Tony Rabbit in Tutorials
EIP 4844, also called proto-danksharding, introduced blobs to Ethereum and made Layer 2 transactions far cheaper. Here is how blobs and the upgrade work.
EIP 4844, better known as proto-danksharding, is the Ethereum upgrade that gave rollups their own cheap, temporary data lane and slashed Layer 2 transaction fees. It shipped as part of the Dencun upgrade, which activated on Ethereum mainnet on March 13, 2024. If you have ever wondered why swapping tokens on Arbitrum, Optimism, or Base suddenly got so much cheaper, EIP 4844 is the main reason.
What is EIP 4844?
EIP 4844 is an Ethereum Improvement Proposal that introduced a brand new transaction type carrying large bundles of data called blobs. The name proto-danksharding signals that this is an early, partial version of a much bigger scaling plan named danksharding, which is itself named after researcher Dankrad Feist. Proto-danksharding ships the transaction format and the fee market for blobs without yet implementing the full data availability sampling that danksharding will eventually require.
The core idea is simple. Rollups such as Arbitrum, Optimism, Base, and zkSync post their compressed transaction data back to Ethereum so the base layer can guarantee that data is available. Before EIP 4844, they had to use regular calldata, which is stored permanently and is expensive. EIP 4844 created a cheaper, dedicated space instead.
What are blobs and how do they work?
A blob is a large packet of data attached to a special blob-carrying transaction (type 0x03). Blobs are not accessible to the Ethereum Virtual Machine, so smart contracts cannot read their raw contents. They exist purely to make data available for a limited time. The key innovation is that blobs are pruned, meaning consensus nodes only keep them for a minimum retention window of roughly 18 days (4096 epochs) before they can be deleted.
This temporary storage is exactly what rollups need. A rollup must publish its data so anyone can reconstruct the chain and challenge fraud, but it does not need Ethereum to store that data forever. By separating temporary blob space from permanent calldata, EIP 4844 made posting data dramatically cheaper. The retention window is long enough for anyone to download the data, verify it, and dispute an invalid state if needed, while still freeing nodes from storing it permanently.
- Separate fee market: Blobs have their own gas pricing that adjusts based on blob demand, so blob fees do not directly compete with normal Ethereum gas.
- Fixed-size data: Each blob is a fixed chunk of data, making capacity easy to reason about per block.
- Verified once, pruned later: Nodes verify blob data, then can discard it after the retention window, keeping long-term storage costs low.
Why did Layer 2 fees drop?
When the Dencun upgrade went live in March 2024, the effect on rollups was immediate. Median fees on major Layer 2 networks fell sharply, in many cases from tens of cents per transaction down to a fraction of a cent. The reason is structural rather than temporary: rollups stopped paying for permanent calldata and started using the far cheaper blob space.
Because the data cost is one of the largest components of a rollup transaction, cutting it down cascaded into much lower fees for end users. This lower-cost environment helped rollups grow into the dominant share of Ethereum activity, and it made everyday actions like swaps, transfers, and minting practical on Layer 2. It also lowered the barrier for new users who were previously priced out of mainnet, and it encouraged more applications to deploy on rollups. If you trade tokens on rollups, you can track DEX activity, liquidity, and new pairs across these networks using DEXTools, which aggregates on-chain data so you can see where volume is actually flowing.
Before and after EIP 4844
The table below summarizes how data posting and fees changed for rollups once blobs arrived.
| Aspect | Before EIP 4844 | After EIP 4844 |
|---|---|---|
| Data storage method | Permanent calldata | Temporary blobs |
| Retention | Stored on chain indefinitely | Pruned after about 18 days |
| Fee market | Shared with all gas | Separate blob fee market |
| Typical L2 user fee | Often tens of cents | Often a fraction of a cent |
| EVM access to data | Readable by contracts | Not readable by the EVM |
Proto-danksharding vs full danksharding
Proto-danksharding is a stepping stone, not the final destination. It introduced the blob format and fee market so the ecosystem could start using blobs right away. Full danksharding is the long-term goal that adds data availability sampling (DAS), a cryptographic technique that lets nodes verify that blob data is available by checking only small random samples instead of downloading everything. That sampling is what allows the network to safely support far more blobs per block.
| Feature | Proto-danksharding (EIP 4844) | Full danksharding |
|---|---|---|
| Blob format and fee market | Yes | Yes |
| Data availability sampling | No | Yes |
| Blobs per block | Started at a target of 3, max 6 | Targets up to 64 blobs |
| Role | Early stepping stone | Full scaling vision |
What came next on the roadmap
EIP 4844 was the start of a phased scaling journey, and the network has kept building on it.
- Pectra (May 7, 2025): This upgrade raised blob capacity through EIP 7691, moving the per-block blob target from 3 to 6 and the maximum from 6 to 9. More blobs means more room for rollups and continued low fees.
- Fusaka (December 3, 2025): This upgrade introduced PeerDAS (Peer Data Availability Sampling, defined in EIP 7594). PeerDAS lets nodes sample blob data instead of downloading all of it, which reduces the bandwidth each validator needs and unlocks safely scaling blob space further. Importantly, Fusaka delivered PeerDAS as a major step toward full danksharding, but it is not full danksharding itself.
In short, as of 2026 the situation is clear: proto-danksharding (EIP 4844) shipped in 2024, Pectra increased blob limits in 2025, and Fusaka added PeerDAS in late 2025. Full danksharding, with its complete data availability sampling and much larger blob counts, remains a planned future milestone rather than a feature that has already launched.
What this means for traders
Cheaper Layer 2 fees changed how people use crypto day to day. Low-cost swaps and transfers made rollups attractive for active trading, arbitrage, and small transactions that were once too expensive on Ethereum mainnet. As more DEX volume migrates to rollups, having a reliable way to monitor pairs and liquidity matters. You can use DEXTools to track Layer 2 tokens, watch new listings, and review on-chain trading activity across multiple rollup networks from one place.
This article is educational and is not financial, investment, or tax advice. Always do your own research before interacting with any token or protocol.
EIP 4844 quietly reshaped Ethereum by giving rollups a purpose-built, temporary data lane. Proto-danksharding lowered Layer 2 fees, validated the rollup-centric roadmap, and laid the groundwork for the data availability sampling that full danksharding will one day complete. Understanding blobs is now essential for anyone following Ethereum scaling in 2026.
Beyond the Blob: The L2 Developer's Perspective on Proto-Danksharding
While users primarily experience the cost savings, for Layer 2 developers, EIP-4844 represents a fundamental shift in how data availability is managed and priced. Before blobs, L2s relied on calldata, which was permanently stored on the Ethereum blockchain. This permanence, while secure, came at a premium, as storage is significantly more expensive than temporary data availability.
Proto-Danksharding's introduction of blobs offers a dedicated, cheaper, and ephemeral data availability layer tailored for rollup needs. This isn't just a discount; it's a re-architecting of the cost model, allowing L2s to scale their transaction throughput without incurring prohibitive data costs. The ephemeral nature of blobs perfectly aligns with the rollup's responsibility to eventually settle a succinct proof on the mainnet, rather than storing every individual transaction.
Practical Implications for L2 Development Teams
- Optimized Data Batching: Developers must strategically batch transactions to fill blobs efficiently, maximizing cost savings per unit of data.
- Transition from Calldata: Existing L2s had to refactor their data publishing mechanisms to utilize blobs instead of calldata, requiring careful testing and deployment.
- Monitoring Blob Utilization: Understanding blob availability and pricing fluctuations becomes crucial for dynamic fee estimation and operational planning.
- Future-Proofing for Full Danksharding: Proto-danksharding provides a valuable proving ground, allowing L2 teams to build expertise that will be directly transferable to the eventual full danksharding implementation.
- Enhanced Throughput Potential: The increased data availability budget directly translates to higher theoretical transaction limits for rollups.
Frequently Asked Questions
What is EIP-4844?
EIP-4844, also called proto-danksharding, is an Ethereum upgrade that introduced a new transaction type carrying data blobs. It was designed as a step toward cheaper data availability for Layer 2 networks.
What are blobs in Ethereum?
Blobs are large packets of data that can be attached to transactions but are not permanently stored by the chain like regular calldata. They give rollups a cheaper way to post their data to Ethereum.
How does proto-danksharding help Layer 2 fees?
By giving rollups a dedicated and cheaper space for posting data, proto-danksharding reduces the cost of submitting Layer 2 transaction data to Ethereum. This can lower fees for users of those rollups.
Why are blobs temporary?
Blob data is only retained by the network for a limited period rather than forever, which keeps storage requirements lower. This temporary availability is enough for rollups to prove and verify their data.