Ethereum gas fees drop to fractions of a cent by 2026

Ethereum transaction costs have fallen sharply through 2026, with average mainnet gas prices sitting around 0.16 gwei as of the year’s data, translating into simple transfer fees below one cent under typical network conditions. The decline follows a string of protocol upgrades, most recently December 2025’s Fusaka fork, that have systematically reduced the cost of both base-layer transactions and the data that Layer 2 rollups post back to Ethereum mainnet.

How gas pricing works on Ethereum

Every action on Ethereum, from a simple ETH transfer to a complex DeFi swap, consumes a measured amount of computational and storage resources, denominated in units of gas. Each type of operation has a fixed gas cost written into the protocol, while the price paid per unit of gas floats according to how much demand there is for block space at any given moment, similar to how a toll road might charge more during rush hour. A simple ETH transfer costs a fixed 21,000 gas units; the total fee paid is that fixed amount multiplied by the current gas price, which is itself set by an auction-like mechanism where users effectively bid for inclusion in the next block. When network demand is low, gas prices fall toward the network’s baseline cost of processing a transaction; when many users compete for limited block space at once, prices rise sharply, which is why gas fees have historically spiked during major NFT launches or periods of market volatility.

How blob transactions changed the cost structure

The reduction traces back to EIP-4844, the “proto-danksharding” upgrade delivered in the Dencun hard fork, which introduced a temporary, cheaper data structure called blobs specifically for Layer 2 rollups to post their transaction data to Ethereum. That change alone cut Layer 2 data costs by as much as 90% in many scenarios compared to the prior method of posting data as regular calldata. Fusaka built directly on that foundation: its PeerDAS mechanism lets validators sample small portions of blob data rather than downloading it in full, cutting validator data loads by roughly 85% and allowing the network to safely support several times more blob capacity than before, which further lowered the per-transaction data cost that Layer 2s pass on to end users.

Blobs work as a separate fee market from ordinary transaction gas, with their own supply of block space and their own floating price. Before Dencun, Layer 2 rollups posted their transaction data to Ethereum as regular calldata, competing directly with every other type of transaction for the same limited block space and paying the same gas prices as any other activity on the network. By carving out a dedicated, temporary data lane specifically for rollups, blobs let Layer 2 data costs float independently of, and generally far below, the price of ordinary mainnet activity, which is the structural reason Layer 2 fees have fallen so much further than base-layer fees over the same period.

The role of the fee-burning mechanism

Since Ethereum’s London upgrade in 2021, a portion of every transaction fee, the base fee, is destroyed rather than paid to validators, a mechanism designed to make ETH’s supply responsive to network demand rather than only to its fixed issuance schedule. Only a smaller “priority fee,” or tip, on top of the base fee goes to the validator that includes the transaction. As base fees have fallen through 2026 alongside the broader decline in gas prices, the amount of ETH burned per transaction has fallen correspondingly, meaning the network now needs substantially higher transaction volume to burn the same aggregate amount of ETH it once burned at lower volumes with higher per-transaction fees. This is part of why some analysts have shifted their attention from gas price alone toward the combination of transaction volume and fee levels when assessing Ethereum’s net issuance trend.

Base layer capacity also increased

Beyond blob-specific changes, Fusaka allowed core developers to raise Ethereum’s default Layer 1 gas limit to 60 million, doubling the network’s underlying base-layer transaction throughput. Combined across Dencun, Pectra and Fusaka, the three upgrades have progressively lowered transaction costs by changing how the network stores and processes the data that rollups depend on, according to technical summaries of the changes from Ethereum infrastructure providers.

Where fees stand now

Current estimates put typical mainnet fees for simple operations in the range of $0.10 to $0.20 under normal conditions, while activity on Layer 2 networks such as Arbitrum, Base and Optimism, which benefit most directly from cheaper blob data, can cost as little as a fraction of a cent to five cents per transaction. Fees still spike during periods of network congestion, such as high-demand NFT mints or major DeFi liquidation events, but the baseline cost floor has moved meaningfully lower than in prior years.

The trade-off behind cheaper fees

Lower fees are not a free lunch from a network-design perspective. Ethereum’s core developers have historically balanced two competing goals: keeping the cost of using the network low enough for ordinary users, and keeping the cost of running a validating node low enough that participation remains broadly accessible rather than concentrated among well-resourced operators with expensive hardware. PeerDAS is notable precisely because it improves both sides of that trade-off at once, letting the network support more blob capacity, and therefore lower per-transaction costs, without proportionally raising the bandwidth and storage burden on individual nodes. Earlier scaling approaches that simply raised the gas limit or blob count without a mechanism like data availability sampling would have pushed node operating costs up in tandem with network capacity, which is why Fusaka’s approach is viewed by Ethereum’s developer community as a more sustainable path to lower fees than brute-force capacity increases alone.

Why it matters for ETH holders

Lower fees do not directly change ETH’s price, but they do lower the barrier to using Ethereum and its Layer 2 ecosystem for everyday transactions, which supports the long-term thesis that network usage, and the ETH burned or paid as fees, can grow as a share of overall crypto activity. For Canadian users and investors, cheaper gas fees make direct on-chain activity, from DeFi to self-custody transfers, more practical relative to earlier years when even simple transactions could cost several dollars during periods of congestion. It is worth noting that lower per-transaction fees also mean lower absolute ETH burned per transaction under Ethereum’s fee-burning mechanism introduced in 2021, so the network’s deflationary pressure from transaction fees now depends more heavily on overall transaction volume than on the size of any individual fee.