Ethereum's Glamsterdam Upgrade: What It Means for Your Gas Costs
Glamsterdam targets roughly 3x L1 capacity and 10x throughput on Ethereum. Here's what the three core EIPs change and what it means for swap costs.
Ethereum is preparing its most structurally significant upgrade since The Merge. Glamsterdam reached its final devnet phase in June 2026 and is now targeted for the second half of 2026, changing how the network builds and executes blocks at the protocol level — and the downstream effect is a network that can absorb far more activity before gas fees spike, across everything from simple ETH transfers to multi-step DeFi operations.
If you execute swaps on Ethereum mainnet, this directly affects what you pay. And unlike timing tricks like swapping at 3am UTC to catch a quiet period, this is a permanent structural reduction — baked into the protocol, not contingent on network conditions.
What Glamsterdam Is Changing at the Core
Glamsterdam bundles several proposals, but three changes account for almost all of the user-facing impact.
1. EIP-7732 — Enshrined Proposer-Builder Separation (ePBS)
Today, when Ethereum blocks are built, validators rely on a network of off-chain relays and external block builders to assemble the most profitable blocks. This system (called MEV-Boost) has worked since The Merge, but it operates entirely outside the protocol, relying on trusted intermediaries rather than enforced rules.
ePBS moves block building into the protocol itself. Validators remain responsible for proposing blocks, but a credentialed builder class handles the assembly. The rules governing this interaction are encoded in consensus — not enforced by trust.
This matters for swap users because MEV is a direct tax on your trades. Sandwiching, front-running, and back-running extract value from your transactions before they settle. Enshrining PBS doesn't eliminate MEV, but it creates a cleaner, more auditable environment where fairness rules can be enforced at the protocol level rather than hoped for at the relay level.
2. EIP-7928 — Block-Level Access Lists (BALs)
Current Ethereum execution is sequential: transactions process one after another within a block, and clients can't know in advance which storage slots each transaction will touch. Block-Level Access Lists change this by requiring each block to declare upfront which accounts and storage slots it will access.
With that map available before execution starts, clients can preload the relevant state and run independent transactions in parallel. Sequential execution is one of the main bottlenecks limiting Ethereum's throughput today. BALs enable parallel processing without a full EVM redesign, which makes them one of the most practically significant EIPs the network has shipped in years.
3. EIP-7904 — Gas Repricing Research
Gas costs on Ethereum are measured in units tied to computational effort, and researchers spent much of early 2026 benchmarking whether that calibration had drifted from what operations actually cost on modern, parallel-execution-optimized clients. EIP-7904 documents that analysis.
The conclusion, finalized in June 2026: no repricing is needed. Once BALs enable parallel execution, current clients already hit the target throughput at existing opcode costs, so EIP-7904 shipped as an informational EIP with no changes to the gas schedule. That changes the mechanism behind Glamsterdam's cost impact — savings come from fitting more transactions into each block at today's per-unit gas costs, not from opcodes getting individually cheaper. A Uniswap trade that currently costs $3–8 in gas fees still gets cheaper on average, mainly because the congestion-driven premium on top of the base cost becomes rarer.
How the Changes Stack Up
BALs and the gas limit increase compound each other — parallel execution keeps block propagation safe at higher capacity, and more transactions fit in each block at today's per-unit gas cost, which means the effective throughput gain is larger than either change alone.
| Change | What It Does | User Impact |
|---|---|---|
| EIP-7732 (ePBS) | Block building moved into protocol | Fairer MEV environment, reduced sandwich risk |
| EIP-7928 (BALs) | Parallel execution via state maps | Higher throughput, fewer congestion spikes |
| EIP-7904 (Gas Repricing) | Concluded no repricing needed | No per-opcode discount; unit costs unchanged |
| Gas limit increase | 60M → 200M gas per block | ~10x throughput, targeting ~10,000 TPS |
The gas limit increase from 60 million to 200 million gas per block is a separate parameter change, but it depends on BALs to work safely. Without declared access lists, tripling the gas limit would make block propagation dangerously slow. The proposals are designed to ship together.
What This Means If You Swap on Ethereum Mainnet
Three things change for the average swap user.
First, congestion-driven premiums shrink. Today, the gas fee on a standard Ethereum swap varies with congestion in real time. Per-unit gas costs stay roughly where they are today — the EIP-7904 research concluded opcode repricing wasn't needed — but with roughly 3x more capacity per block, moderate traffic is far less likely to push you into the premium you'd pay today. Quiet-period gas savings still stack on top of that.
Second, congestion events get less severe. The combination of a tripled gas limit and parallel execution means the network absorbs more activity before it hits its ceiling. Fee spikes during volatile market periods — the moments when you most want to swap and least want to pay $40 in gas — will still happen, but they start from a lower floor and take longer to develop.
Third, multi-step DeFi becomes more competitive with L2s. Today, complex operations like a routed swap touching multiple liquidity pools might cost 80–95% less on Arbitrum or Base than on mainnet. After Glamsterdam, that gap narrows. Some DeFi activity may shift back to L1, which matters if you prioritize mainnet security guarantees over the trust assumptions that come with rollups.
That doesn't make L2s obsolete. For frequent low-value swaps, Arbitrum, Base, and other rollups will still win on cost. But the calculus shifts for larger transactions where the mainnet security model is worth more.
If you want to understand how current gas pricing works before Glamsterdam lands, this breakdown of gas fees and congestion covers the mechanics and how to time swaps around them.
Timeline and What Could Slip
Glamsterdam reached its final devnet phase in June 2026, and the target has shifted from the original H1 2026 window to the second half of the year. Public testnet seasoning on Holesky and Hoodi comes next, and mainnet activation realistically lands between September and December 2026 — the interplay between ePBS and BALs introduces complexity that hasn't been tested at mainnet scale.
Things worth watching:
- Devnet and testnet results — If ePBS and BAL interactions surface bugs under load, the timeline moves. Ethereum's execution cadence improved with Pectra and Fusaka, but Glamsterdam's scope is larger than both combined.
- Gas repricing, resolved rather than delayed — After benchmarking against parallel-execution-optimized clients, EIP-7904 concluded no opcode repricing is needed to hit the target throughput; it shipped as an informational EIP in June 2026 with no client changes required. That removes one item from the launch-blocking list, though the smaller gas-and-state repricing bundle tracked under EIP-8007 is still being finalized.
- Builder ecosystem readiness — ePBS requires a credentialed builder market operational on day one of mainnet. If tooling or infrastructure isn't ready, block quality during the transition period could suffer.
- Hegotá dependency — Ethereum's roadmap positions Glamsterdam as a prerequisite for the following fork, Hegotá, which is expected to include post-quantum account abstraction via EIP-8141 — currently under "considered for inclusion" review rather than a committed headliner. A Glamsterdam slip pushes the entire downstream roadmap.
The realistic framing: with public testnets targeted for the second half of 2026, mainnet activation most likely lands in Q4 2026, pushing the structural gas cost impact to late this year at the earliest. A further slip into 2027 remains possible if ePBS testing surfaces issues.
Use the Ethereum Gas Tracker to monitor current mainnet conditions in real time while you wait.
The Bottom Line
Glamsterdam is a protocol-level restructuring, not a market event. The gas cost improvement comes from tripling block capacity and enabling parallel execution, not from a blanket opcode discount — research into broader gas repricing (EIP-7904) concluded in June 2026 that current costs already meet the throughput target. The upgrade's effect still lands the moment it activates, regardless of demand: when it ships, every swap on Ethereum mainnet becomes far less likely to hit a congestion-driven fee spike.
The main risk is timeline. If gas costs are material to your current swaps, the L2 Gas Tracker can show you cheaper execution alternatives on Arbitrum, Base, and other rollups in the meantime. ETH is available to swap on Zest Exchange.