TL;DR
MegaETH is an Ethereum Layer 2 designed for real-time execution: sub-10ms block times, 100,000+ transactions per second, and 10+ gigagas per second of execution throughput, while settling to Ethereum for security.
Mainnet launched in early 2026. The MEGA token generation event took place April 30, 2026.
The architectural breakthrough is SALT (Small Authentication Large Trie), which holds the authentication structure in RAM and removes the disk I/O bottleneck that has limited every blockchain since Bitcoin.
The chain targets high-frequency DeFi, real-time onchain games, and consumer apps that need web2-level responsiveness without leaving the EVM.
Across supports MegaETH on mainnet (chain ID 4326) from day one. Bridge to MegaETH.
MegaETH stops pretending block times are a constraint
Every EVM chain on the market has accepted some version of the same trade-off: blocks land every few hundred milliseconds at fastest, finality lags by seconds or minutes, and the bottleneck is roughly the same place. Disk I/O for moving state in and out of storage. Speed up consensus and you still wait on storage. Move to a faster validator set and you still wait on storage. The hardware constraint underneath the software was the same.
MegaETH rebuilt that part of the stack. The team's bet is that the authentication structure for state, the Merkle trie that proves which accounts exist and how much they hold, can live entirely in RAM instead of on disk. The result is the closest thing the EVM ecosystem has shipped to a real-time blockchain: 10-millisecond block times, throughput targets above 100,000 TPS, and execution that feels less like batch settlement and more like a live system.
Mainnet is live. The token, MEGA, generated April 30, 2026 against verified on-chain KPIs. Across supports the chain on day one. Below is the actual architecture, what it changes for builders and users, and how to move funds onto it.
What MegaETH actually is, technically
MegaETH is an Ethereum Layer 2 rollup that settles to Ethereum L1, which means it inherits Ethereum's economic security for finality the same way Arbitrum, Optimism, Base, and Unichain do. What's different is the execution layer running on top.
The performance targets the team is building toward:
Sub-10ms block times in steady-state operation
100,000+ transactions per second sustained throughput
10+ gigagas per second of execution throughput
Full EVM compatibility, so existing Solidity and tooling work without modification
Ethereum-anchored settlement, so withdrawals and bridge security inherit L1's guarantees
Chain ID is 4326. Native asset for gas is ETH. Bridged stablecoin support at launch is USDT0. The application surface is identical to any other EVM rollup: MetaMask, Rabby, hardware wallets, Wagmi/Viem on the dev side, Foundry/Hardhat for contracts.
SALT is the load-bearing innovation
The piece of MegaETH that's genuinely new is SALT, Small Authentication Large Trie. Every blockchain has to prove, on every transaction, that the state it's reading and writing actually exists and is authenticated by the chain's history. That proof comes from a Merkle trie, and every node lookup has historically meant a disk read. Disk reads are slow. They've been the throughput ceiling for every L1 and L2 the EVM has produced.
SALT keeps the authentication structure in memory while leaving the bulk state on disk. The trie that proves "this account exists, this balance is canonical" never touches disk during transaction processing. The disk only stores the underlying data, not the authentication path to it.
The practical effect is that MegaETH can execute orders of magnitude more transactions per second without the I/O bottleneck that has constrained every previous EVM chain. The architecture is closer to how a database with heavy read caching operates than to how most blockchains move state around.
What the speed actually enables
A chain with 10ms blocks isn't faster Ethereum. It's a different category of application surface. A few workloads matter here.
High-frequency DeFi. Onchain orderbooks, perpetual exchanges, and structured products that need to update prices and execute liquidations on a sub-second timescale. On most EVMs, these run as off-chain matching engines with on-chain settlement because the chain itself is too slow to host the matching. On MegaETH, the matching can be onchain. Aave V3 and Chainlink are among the protocols that have shipped to the network. Cap Money has surpassed $400M TVL on stress test conditions according to public benchmarks.
Real-time onchain games. Fully onchain game state, where every move and every interaction touches the chain, is impractical on chains where each turn costs gas equivalent to a Ferrari per hour. On MegaETH, the throughput and gas profile make fully onchain games tractable. Several titles, including Biomes (a Minecraft-style sandbox) and Stomp (a PvP monster battler), are using the chain for exactly this.
Consumer applications. Apps that need to feel responsive (social, payments, creator tools) have historically had to fake it with off-chain UX layered on top of slow onchain settlement. MegaETH lets the chain itself respond at speeds that feel like a normal app. Euphoria (mobile-first derivatives), Lemonade (creator OS), and others are building on this premise.
How MegaETH fits into the Ethereum scaling roadmap
MegaETH isn't trying to replace Ethereum and isn't trying to compete with the broader L2 ecosystem on cost. It's competing on one specific axis: latency. The chain settles to Ethereum like any other rollup, inherits L1's security like any other rollup, and runs the EVM like any other rollup. What it does differently is collapse the time between when a user signs a transaction and when that transaction is executed and observable, by roughly two orders of magnitude.
For the Ethereum roadmap, MegaETH plays a specific role: it is the chain where applications that need real-time execution can stay onchain instead of running matching off-chain and settling on-chain. That's a different value proposition than cheaper-than-Ethereum. It's executable in a way Ethereum L1 fundamentally is not.
The teams building on it tend to be teams that genuinely need the latency. Market makers. Game studios. Consumer apps with millisecond-sensitive UX loops. The applications that don't need real-time execution will continue to live on cheaper rollups optimized for different workloads.
How to get onto MegaETH
Across supports MegaETH bridging on mainnet from day one. The route covers ETH and USDT0 from over 22 source chains, including Ethereum, Solana, BNB Smart Chain, Arbitrum, Base, Optimism, and the rest of the production network.
Open the Across dapp (https://across.to/megaeth-bridge?utm_source=blog&utm_medium=blog&utm_campaign=what-is-megaeth-20260527) and connect your wallet. Select your origin chain and asset, set MegaETH as the destination, review the quote, and confirm. A relayer fills the intent on MegaETH from their own capital, you receive canonical ETH or canonical USDT0 in your wallet, and the protocol settles with the relayer in the background through UMA's Optimistic Oracle.
The bridge takes roughly two seconds for L2-to-MegaETH transfers under $10,000. There's no wrapped asset in your wallet. When you bridge ETH, you receive ETH. There's no locked liquidity pool on MegaETH to attract attackers. Across has processed over $35 billion in crosschain volume for more than 2.5 million users with zero protocol-level security exploits since launch in 2021.
What builders should know if they're shipping to MegaETH
If you're a developer, three things make MegaETH a low-friction target.
First, EVM compatibility is full. Your existing Solidity, your existing audit reports, your existing CI pipeline, your existing infrastructure tooling: none of it needs to change. You deploy to MegaETH the same way you deploy to Ethereum or any rollup.
Second, the latency profile changes which applications are economically viable. Designs that didn't pencil out on Ethereum L1 (frequent state updates, sub-second oracle reads, per-action onchain logic) work on MegaETH. If you have a design idea that was previously "too expensive" or "too slow," it's worth re-evaluating against this chain.
Third, the bridge story matters for any multichain app. Across exposes a single API and a single integration model across every chain it supports, so adding MegaETH to a multichain product is the same integration shape as adding Base or Arbitrum.
Get onto the chain that stopped accepting the disk bottleneck
MegaETH is what happens when a team decides storage I/O is a software problem, not a physics problem, and rebuilds the bottleneck. The chain is live, the applications that benefit most are already shipping, and the bridge is ready when you are.