Ethereum’s September 2026 gas fee spike: five myths about blobs

Fee fluctuations and rollup demand

The Ethereum network experiences fee volatility because real-time demand for blockspace remains the primary driver of transaction costs. Trading memecoins, issuing new tokens, and high-frequency DeFi applications cause gas prices to rise for both mainnet and rollups. While EIP-4844 introduced blob transactions to lower Layer 2 operating costs, users still face blob market congestion during synchronized periods of rollup demand.

Current data shows that rollups like Arbitrum, Optimism, and Base account for a massive share of retail applications. These networks process transactions for cents, but they still rely on Ethereum for settlement and data availability. This dependency creates a multi-layer fee model where congestion on one layer affects the others.

Transaction Type	Typical 2026 Mainnet Cost
ETH Transfer	$0.10 to $0.25
ERC-20 Transfer	~$0.076
Swap	~$0.21
NFT Minting	Significant reduction vs 2021

The difference between 2021 and 2026 is massive. During the 2021 boom, NFT mints often cost over $100, whereas modern rollups execute these tasks for fractions of a cent. You might notice that L2 transaction costs depend on the specific architecture of the rollup used. ZK-rollups prioritize compression and throughput, while optimistic rollups focus on liquidity within the network.

Addressing blob fee myths

The claim that blobs pay zero transaction fees is false. Submitters pay between $0.10 and $3.00 in mainnet execution fees per blob depending on the data type and gas prices. Most of the current fee distribution stays concentrated in execution fees rather than blobspace fees.

Another common misconception suggests that the protocol earns nothing from blobs. While direct blob burn remains small, associated execution fees are material. Some researchers propose a higher reserve to increase burn, but evidence shows this could reduce total fee capture by lowering usage. Do not assume that increasing the reserve alone compensates for operator costs.

The implementation of EIP-7762 introduces a minimum blob base fee of roughly $0.01. This change aims to reduce the time the market spends in priority gas auctions during demand surges. On June 20, it took six hours of excess demand for the network to reach equilibrium following a surge. The $0.01 fee should minimize future dislocations without significantly raising total costs for users.

Different entities use different submission strategies. Base and OP Mainnet post data to an EOA, which requires only 21,000 gas regardless of the number of blobs. This efficiency allows them to carry five or more blobs at very low costs. In contrast, Arbitrum and ZkSync Era post more complex data and often submit only a single blob per transaction.

MEV-burn and auction mechanics

The MEV-burn auction pilot examines how to incentivize earlier bidding to deter builder collusion. In a scenario where builders collude with proposers, a proposer could rebate a large portion of a tip back to a builder. If the proposer wants to keep a 0.2 ETH tip, they could rebate up to 1 ETH to builders. This mechanism addresses the risk where builders set base fee floors to manipulate selection.

One proposal suggests a bid validity condition where a bid at the block cannot exceed the base fee floor by more than 15 percent. This adds complexity to the auction process. Current research into smoothing the blob reference also shows tradeoffs. A two-hour candidate substitute lowers the p99.9 price by 50.7 percent, but it also raises the median price by 17.4 percent.

The network must balance throughput with the cost of serving blobs for home stakers. Increasing the target and maximum blob capacity provides burst headroom, but it does not guarantee stable prices during every burst. Developers continue to use batching and compression to keep these costs low. Can the network maintain this stability as demand grows toward the next capacity increase?