Bitcoin vs Ethereum Transaction Fee Comparison

Bitcoin vs Ethereum Fees Overview

Bitcoin and Ethereum use fundamentally different fee models. Bitcoin charges per byte of transaction data using a sats/vByte model, where fees depend on transaction size rather than value. Ethereum charges per unit of computation using a gas/gwei model governed by EIP-1559, where fees depend on the complexity of the operation being performed. Both networks have seen dramatic fee reductions in 2026: Bitcoin averages around $0.30 per transaction (down 79% year-over-year), while Ethereum sits at roughly 0.43 gwei (down 35% year-over-year), both near historic lows.

The following table compares typical costs for common operations on each network and their respective Layer 2 solutions, using live data from June 2026.

Note: These costs reflect June 2026 fee levels, which are historically low on both networks. During congestion events, fees can spike 100x or more. Always check current rates using a fee estimator before transacting.

How Bitcoin Fees Work: Sats/vByte Model

Bitcoin fees are determined by transaction size, not transaction value. Sending $10 and $10 million costs the same if the transactions have identical byte sizes. Fee rates are denominated in satoshis per virtual byte (sat/vB), and miners prioritize transactions offering higher sat/vB rates.

The virtual byte system was introduced with SegWit to provide a discount on witness (signature) data. Non-witness data counts as 4 weight units per byte, while witness data counts as 1 weight unit per byte. Virtual bytes equal total weight divided by 4. This discount incentivizes SegWit adoption and makes P2WPKH transactions roughly 38% cheaper than legacy P2PKH transactions.

Taproot (P2TR) transactions push savings further: a key-path spend input is approximately 57.5 vB compared to 68 vB for a P2WPKH input, a 15% reduction. For multisig wallets, the difference is dramatic: a legacy 2-of-3 multisig input costs about 296 vB, a P2WSH input costs 104 vB (65% savings), and a MuSig2/FROST Taproot multisig collapses to a single key-path spend at 57.5 vB (81% savings over legacy).

Bitcoin blocks are limited to 4 million weight units (roughly 1 to 1.5 MB of actual data after SegWit), so fee rates rise when demand for block space exceeds supply. For a deeper look at the economics of Bitcoin block space, see our research on Bitcoin fee market dynamics.

How Ethereum Fees Work: Gas and EIP-1559

Ethereum fees are determined by computational complexity. Each operation in the Ethereum Virtual Machine consumes a fixed number of gas units: a simple ETH transfer always costs 21,000 gas, an ERC-20 token transfer costs 45,000 to 65,000 gas, and a Uniswap V3 swap costs 150,000 to 185,000 gas. The total fee equals gas units consumed multiplied by the effective gas price in gwei (1 gwei = 0.000000001 ETH).

Since August 2021, Ethereum has used the EIP-1559 fee mechanism, which splits the gas price into two components: a base fee set algorithmically by the protocol and a priority fee (tip) set by the user. The base fee adjusts up or down by a maximum of 12.5% per block depending on whether blocks are above or below 50% capacity. Users set a maxFeePerGas (ceiling they are willing to pay) and a maxPriorityFeePerGas (tip to validators). The base fee portion is burned permanently, removing ETH from circulation.

This mechanism makes gas prices more predictable than Bitcoin's pure auction model. During low congestion (like June 2026), the base fee can drop below 0.5 gwei, making even complex DeFi operations cost under $0.30. During congestion spikes, the base fee rises rapidly to suppress demand.

Historical Fee Spikes

Both networks have experienced extreme fee events that demonstrate how quickly costs can escalate when demand outstrips block space.

Bitcoin Fee Spikes

The most significant Bitcoin fee spike in history occurred on April 20, 2024, when the Runes protocol launched simultaneously with the fourth halving. Average fees hit $91.89 per transaction, a 2,645% increase from $3.35 one month earlier. A single block generated over $2.4 million in fees, the highest-fee block ever mined. Runes consumed more than 90% of all network fees in the days following launch.

Earlier, in May 2023, the Ordinals and BRC-20 craze pushed average fees above $19, with the mempool swelling beyond 500 MB. On May 7, 2023, transaction fees exceeded the block subsidy for the first time since December 2017.

By 2025, both Runes and Ordinals activity had cooled dramatically. Runes dropped from 90% of daily transactions to under 2%, and average fees settled to around $1.46. In June 2026, fees sit at historic lows of $0.24 to $0.40, with all priority tiers at 1 sat/vB.

Ethereum Fee Spikes

Ethereum's most sustained fee crisis ran from mid-2021 through 2022, when NFT marketplace activity on OpenSea pushed average transaction costs above $100. Gas wars during popular mints saw users bidding hundreds of gwei for priority inclusion.

A sharp spike hit on February 9, 2024, when the experimental ERC-404 token standard drove speculative activity that pushed average gas prices to 70 gwei with spikes to 377 gwei. Another surge in September 2024 saw a 164% monthly increase in gas prices driven by DeFi activity.

By mid-2025, Ethereum mainnet fees had plunged to five-year lows, averaging just $0.168 per transaction, as activity migrated to L2 networks. June 2026 continues this trend with gas prices under 0.5 gwei and typical swap costs below $0.30.

Layer 2 Fee Comparison

Both ecosystems have developed Layer 2 solutions that reduce fees by orders of magnitude. The approaches differ: Bitcoin L2s focus on payment channels and statechains, while Ethereum L2s use rollups that inherit mainnet security.

Ethereum L2 fees have dropped 10 to 100x since the Dencun upgrade (March 2024) introduced blob transactions via EIP-4844. Before Dencun, L2s spent roughly $34 million per month on calldata. After the upgrade, total blob fees for the entire first year came to approximately 3,000 ETH (~$8 million). The Pectra upgrade (May 2025) doubled blob capacity, and the Fusaka upgrade (December 2025) introduced PeerDAS, enabling up to 8x further scaling. For a broader view of fees across chains, see our chain fee comparison tool.

Fee Estimation Strategies

The two networks require different approaches to fee estimation and optimization. Understanding these mechanisms can save significant costs, especially during congestion events.

Bitcoin: RBF and CPFP

Bitcoin offers two primary mechanisms for managing fees on unconfirmed transactions. Replace-by-Fee (RBF) allows a sender to broadcast a replacement transaction with a higher fee rate while the original is still in the mempool. This is useful for underbidding initially and bumping only if the transaction does not confirm in time. Child-Pays-for-Parent (CPFP) lets the recipient (or sender, via a change output) create a new transaction spending an unconfirmed output with a high enough fee to incentivize miners to confirm both.

Fee estimation on Bitcoin relies on mempool analysis. Tools like mempool.space provide real-time visualizations of pending transactions grouped by fee rate, allowing users to target the lowest rate likely to confirm within their desired timeframe. The mempool.space Accelerator service (launched July 2024) also offers off-chain fee payments to partner mining pools covering approximately 40% of global hashrate.

Ethereum: EIP-1559 Priority Fees

Ethereum's EIP-1559 mechanism automates most of the fee estimation work. The base fee is set by the protocol and adjusts predictably every block. Users only need to choose their priority fee (tip), which determines inclusion speed relative to other transactions. During low congestion like June 2026, a priority fee of 0.001 gwei is sufficient. During high congestion, the priority fee becomes the main differentiator for block inclusion.

Unlike Bitcoin, Ethereum does not support RBF natively, but wallets can replace pending transactions by resubmitting with the same nonce and a higher gas price. This achieves the same effect as RBF but through a different mechanism.

When Bitcoin Fees Are Cheaper

Bitcoin L1 is cheaper for simple value transfers, especially high-value payments. Because fees are based on data size rather than value, sending $100,000 in bitcoin costs the same $0.03 to $0.10 as sending $10. This makes Bitcoin particularly cost-effective for:

• Large-value settlement between institutions
• Cross-border transfers where the fee-to-value ratio matters
• Batch payments consolidating many outputs into one transaction
• UTXO consolidation during low-fee periods

Bitcoin L2s extend this advantage further. Lightning Network routing fees are typically 20 to 150 parts per million (ppm), making micropayments of a few cents economically viable. Spark eliminates transfer fees entirely for bitcoin and USDB movements within the protocol, with users only paying on-chain fees when depositing to or withdrawing from the network.

When Ethereum Fees Are Cheaper

Ethereum's fee advantage appears in programmable transactions where Bitcoin L1 has no equivalent capability. Smart contract interactions, DeFi swaps, lending operations, and complex token logic are all native to Ethereum. For these use cases, comparing to Bitcoin is not meaningful because Bitcoin L1 simply does not support them.

Within its own capabilities, Ethereum L2s offer the lowest fees for token transfers and DeFi operations. A USDC transfer on Base costs roughly $0.01 to $0.02. A swap on Arbitrum costs about $0.27. These costs continue to fall as blob capacity increases.

For a broader comparison of Bitcoin and Ethereum across dimensions beyond fees, see our Bitcoin vs Ethereum comparison.

Fee Trends: 2023 to 2026

Both networks have followed a pattern of fee spikes followed by structural declines, driven by different dynamics.

Bitcoin fees spiked twice (Ordinals in May 2023, Runes at the halving in April 2024) and then declined as speculative token activity faded. Monthly fee revenue fell from $439 million per month in 2024 to $239 million in 2025. The structural driver of lower fees is reduced demand for block space as Runes and Ordinals activity contracted to under 2% of daily transactions.

Ethereum fees have declined for a different reason: successful scaling. The Dencun upgrade (March 2024), Pectra upgrade (May 2025), and Fusaka upgrade (December 2025) progressively increased data availability capacity, enabling L2s to post data at a fraction of their previous cost. This pushed activity off mainnet, reducing L1 demand while making the overall ecosystem cheaper. Ethereum core developers plan to increase blob capacity to 48 blobs per block by mid-2026, with a long-term target of 128 blobs per slot under full Danksharding.

Frequently Asked Questions

Is Bitcoin or Ethereum cheaper to send?

In June 2026, both are extremely cheap for simple transfers. A standard Bitcoin transaction costs roughly $0.03 to $0.10 at 1 sat/vB, while an ETH transfer costs about $0.02 at 0.43 gwei. However, fees are highly variable on both networks. During the April 2024 Runes launch, Bitcoin fees averaged $91.89. During the 2021-2022 NFT boom, Ethereum fees exceeded $100 per transaction. The current low-fee environment is not guaranteed to persist.

Why are Bitcoin fees based on size instead of value?

Bitcoin miners are constrained by block size (4 million weight units), not by the value of transactions they include. A transaction spending $10 and one spending $10 million can have identical byte sizes, so they consume the same block space. Miners maximize revenue by selecting transactions with the highest fee-per-byte, regardless of the value being transferred. This design makes Bitcoin particularly efficient for high-value transfers.

What is the cheapest way to transact on Bitcoin?

The cheapest option is using a Layer 2 solution. Lightning Network routing fees are fractions of a cent. Spark charges no transfer fees for bitcoin and USDB movements, with costs only incurred when moving funds on-chain. For on-chain transactions, using Taproot (P2TR) addresses reduces input sizes by about 15% compared to Native SegWit, and waiting for low-fee periods (typically weekends or early morning UTC) can reduce costs further. You can monitor real-time fee rates at our fee estimator.

How does EIP-1559 affect Ethereum gas fees?

EIP-1559 replaced Ethereum's first-price auction with a predictable base fee mechanism. The base fee adjusts by up to 12.5% per block based on network utilization, making gas costs more predictable. Users set a maximum fee and a priority tip; the difference between the maximum and effective price is refunded. The base fee is burned, which reduces ETH supply when the network is active. This mechanism smooths out sudden fee spikes but does not prevent them entirely during sustained congestion.

How did blob transactions reduce Ethereum L2 fees?

EIP-4844 (deployed March 2024) introduced a new transaction type that carries temporary data blobs instead of permanent calldata. Blobs are stored for approximately 18 days and then pruned, which is sufficient for rollup fraud-proof windows. This reduced L2 data posting costs by 10 to 100x. Before EIP-4844, L2s collectively spent about $34 million per month on calldata. After the upgrade, total blob fees for the entire first year were roughly $8 million. Subsequent upgrades (Pectra and Fusaka) have further increased blob capacity.

Can I speed up a stuck Bitcoin transaction?

Yes, through two mechanisms. Replace-by-Fee (RBF) lets you rebroadcast the same transaction with a higher fee rate, replacing the original in the mempool. Child-Pays-for-Parent (CPFP) lets you spend an unconfirmed output with a high enough fee to make the combined fee rate attractive to miners. RBF requires the original transaction to signal opt-in (most modern wallets do this by default). Our RBF detector tool can check whether a pending transaction is replaceable.

Why are Ethereum L1 fees so low in 2026?

Ethereum L1 fees have dropped to five-year lows because user activity has migrated to L2 networks. The success of rollups like Base, Arbitrum, and Optimism means fewer transactions compete for mainnet block space. At the same time, Ethereum's data availability upgrades (EIP-4844, Pectra, Fusaka) have expanded L2 capacity, further reducing the need for expensive L1 transactions. This is a structural shift rather than a temporary lull: as L2 infrastructure matures, mainnet fees are expected to remain low except during exceptional demand events.

This tool is for informational purposes only and does not constitute financial advice. Fee data is approximate and based on publicly available information from mempool.space, Etherscan, and YCharts as of June 2026. Transaction fees are highly variable and can change rapidly. Always verify current fee rates before transacting.