Glossary

Blob Fee Market

The blob fee market is a separate EIP-4844 fee mechanism that prices data blobs independently from regular Ethereum execution gas.

Key Takeaways

  • The blob fee market is a dedicated pricing mechanism introduced by EIP-4844 (proto-danksharding) that sets blob data costs independently from Ethereum's regular EIP-1559 execution gas market.
  • Blob pricing uses an exponential base fee that adjusts based on demand relative to a target of 6 blobs per block (updated from 3 after the Pectra upgrade), with a minimum floor of 1 wei per blob gas.
  • This separate market dramatically reduced rollup data posting costs: Layer 2 transaction fees dropped by over 90% compared to posting data as permanent calldata.

What Is the Blob Fee Market?

The blob fee market is a separate fee mechanism on Ethereum that prices temporary data blobs independently from regular transaction execution. Introduced by EIP-4844 (also known as proto-danksharding) in the Dencun upgrade on March 13, 2024, it created a new resource dimension on Ethereum with its own supply, demand, and price discovery process.

Before EIP-4844, rollups posted their transaction data as calldata, which is stored permanently on-chain and priced within the same gas market as all other Ethereum execution. This meant rollup data competed directly with DeFi trades, NFT mints, and every other transaction for the same limited block space. The blob fee market solved this by giving rollup data its own lane with its own pricing: blobs are temporary (pruned after approximately 18 days), so they can be priced far cheaper than permanent storage.

How It Works

The blob fee market operates as a parallel fee market alongside Ethereum's execution gas market. Every blob transaction (Type 3) carries two independent fee components: a standard execution gas fee (for the transaction's EVM computation) and a blob gas fee (for the data blobs attached to the transaction).

Target and Maximum Parameters

The blob fee market defines a target and maximum number of blobs per block. When the Dencun upgrade first activated, the parameters were set to a target of 3 blobs and a maximum of 6 blobs per block. The Pectra upgrade (May 7, 2025) doubled throughput via EIP-7691:

ParameterDencun (EIP-4844)Pectra (EIP-7691)
Target blobs per block36
Max blobs per block69
Target blob gas per block393,216786,432
Max blob gas per block786,4321,179,648
Blob base fee update fraction3,338,4775,007,716

Each blob is 128 KiB (4,096 field elements of 32 bytes each) and consumes exactly 131,072 (217) blob gas. Only the KZG commitment to each blob is stored on the execution layer: the blob data itself lives on the consensus layer (beacon chain) and is pruned after approximately 4,096 epochs (about 18 days).

Exponential Base Fee Adjustment

Like EIP-1559 for execution gas, the blob fee market uses a base fee that adjusts automatically based on demand. However, while EIP-1559 applies a linear 12.5% maximum adjustment per block, the blob base fee uses a true exponential function.

The system tracks a running counter called excess_blob_gas. When a block includes more blobs than the target, the counter increases. When it includes fewer, the counter decreases (floored at zero). The blob base fee is then computed as:

blob_base_fee = MIN_BASE_FEE * e^(excess_blob_gas / UPDATE_FRACTION)

Where:
  MIN_BASE_FEE = 1 wei (the absolute floor)
  UPDATE_FRACTION = 5,007,716 (post-Pectra)

To avoid floating-point arithmetic (which would break consensus), the protocol implements this using a deterministic integer-only Taylor series expansion called fake_exponential:

def fake_exponential(factor, numerator, denominator):
    i = 1
    output = 0
    numerator_accum = factor * denominator
    while numerator_accum > 0:
        output += numerator_accum
        numerator_accum = (numerator_accum * numerator) // (denominator * i)
        i += 1
    return output // denominator

The update fraction is calibrated so that a fully saturated block (all 9 blob slots filled) increases the base fee by approximately 8.2%, and an empty block (zero blobs) decreases it by approximately 14.5%. This asymmetry in the post-Pectra parameters (the target-to-max ratio changed from 1:2 to 2:3) means the fee drops faster during low demand than it rises during high demand.

How Blob Transactions Pay Fees

A blob transaction includes a max_fee_per_blob_gas field, similar to EIP-1559's max_fee_per_gas. The transaction is only valid if this value meets or exceeds the current blob base fee. The sender pays the blob base fee (not their maximum bid), and the difference is not charged. This mirrors EIP-1559's design where the base fee is burned rather than paid to validators.

  1. The sender sets max_fee_per_blob_gas as their ceiling bid
  2. The protocol calculates the current blob base fee from excess_blob_gas
  3. If the bid meets or exceeds the base fee, the transaction is valid
  4. The sender is charged: blob_base_fee × blob_gas_used (burned, not paid to validators)
  5. Regular execution gas fees are charged separately through the standard EIP-1559 mechanism

Cost Savings for Rollups

The blob fee market's primary purpose is reducing data availability costs for rollups. Before EIP-4844, rollups paid execution gas rates to post data as calldata: 16 gas per non-zero byte, stored permanently. With blobs, rollups post temporary data at blob gas rates, which are orders of magnitude cheaper during normal demand.

After the Dencun upgrade, the impact was immediate and dramatic:

RollupPre-Dencun CostPost-Dencun CostReduction
Base (Coinbase)~$1.50 per tx~$0.03 per tx~98%
Optimism~$0.50 per tx~$0.001 per tx~99%
Zora$0.10-$0.70 per tx<$0.02 per tx~95%+
zkSync Era0.11 ETH per batch0.013 ETH per batch~88%

Before the upgrade, Layer 2 networks collectively spent roughly $34 million per month on calldata. After blobs, those costs dropped by 90-98%. For a deeper analysis, see the research article on EIP-4844 and the blob fee market.

Use Cases

Optimistic Rollup Data Posting

Optimistic rollups like Optimism and Arbitrum post compressed transaction batches as blobs. The data must remain available long enough for challengers to submit fraud proofs during the challenge period (typically 7 days). Since blobs persist for approximately 18 days, this window is comfortably covered.

ZK Rollup Data Availability

ZK rollups like zkSync Era and Starknet use blobs to publish state differences or transaction data alongside their validity proofs. Even though ZK proofs verify correctness, the underlying data must be available so that users can reconstruct the rollup state and exit if needed.

Data Availability Layer

The blob fee market effectively turns Ethereum into a priced data availability layer. Any protocol that needs temporary data publication with cryptographic commitments can use blobs: not just rollups, but also validiums that want partial on-chain DA, or protocols anchoring data proofs to Ethereum.

Risks and Considerations

Fee Volatility During Demand Spikes

The exponential pricing mechanism means blob fees can spike rapidly when demand exceeds the target. During the "blobscriptions" event in late March 2024, users inscribed data directly into blobs, temporarily saturating capacity and driving blob fees from near-zero to significant levels. Rollups that had not set sufficiently high max_fee_per_blob_gas found their transactions stuck until fees subsided.

Underpriced Resources at the Floor

When blob demand is below the target, the base fee drops to its 1-wei floor, effectively making blob space free. For most of 2024, blob utilization stayed well below the 3-blob target, meaning rollups posted data for negligible cost. While good for rollup users, this reduces Ethereum's fee burn from blobs to near zero, raising questions about long-term security budget sustainability. EIP-7918 has been proposed to address this by bounding the blob base fee relative to execution costs.

Blob Data Expiry

Blobs are pruned after approximately 18 days. Any protocol relying on blob data for longer periods must implement an archival strategy. If a rollup's challenge period exceeds the blob retention window, or if users need historical data for state reconstruction, separate data availability solutions are required.

Competition with Alternative DA Layers

The blob fee market competes with external data availability solutions like Celestia, EigenDA, and Avail. While Ethereum blobs offer the strongest security guarantees (inheriting Ethereum's validator set), external DA layers can offer higher throughput at lower cost. Some rollups use a hybrid approach: posting to Ethereum blobs during low-fee periods and falling back to external DA when blob fees spike.

Implications for Bitcoin Layer 2 Design

The blob fee market offers lessons for Bitcoin scaling. Unlike Ethereum's approach of creating separate data markets, Bitcoin Layer 2 solutions like Spark use off-chain protocols that avoid on-chain data posting entirely. This sidesteps fee market volatility but requires different trust and availability assumptions. Understanding blob fee dynamics helps contextualize the tradeoffs between on-chain data availability (Ethereum rollups) and off-chain state management (Bitcoin L2s).

This glossary entry is for informational purposes only and does not constitute financial or investment advice. Always do your own research before using any protocol or technology.