MegaETH

MegaETH’s first public research framed latency as the missing property of EVM chains. Instead of asking every node to order, execute and verify the same work, it assigned different jobs to sequencers, replicas, full nodes and provers. Only one sequencer would be active in normal operation. That removed consensus overhead from the immediate transaction path and concentrated block production in one high-performance machine.

The bargain was explicit. Replica nodes could remain light because they applied streamed state differences rather than re-executing every transaction; full nodes and provers retained heavier validation roles. MegaETH was not claiming that block production itself was decentralized. It was betting that Ethereum settlement, external data availability and separate validation could contain the trust cost of making execution feel like a server.

The public testnet opened in March 2025 with lightweight mini-blocks arriving around every 10 milliseconds while conventional EVM blocks continued at roughly one-second intervals. Applications could react to the stream without forcing wallets, explorers and contracts to abandon familiar EVM block semantics.

MegaETH also replaced a state-update bottleneck with SALT, an authentication tree designed to fit in memory and avoid random disk work. These choices made the project’s first promise unusually measurable: users could feel a shorter feedback loop, and a later seven-day mainnet stress test processed 11 billion transactions according to the project. Neither figure alone showed that applications would remain on the chain.

In October 2025, MegaETH sold 5% of MEGA through a 72-hour English auction on Sonar, with a $999 million maximum fully diluted valuation. The current token page says more than 50,000 bidders committed $1.39 billion. The MiCA white paper records lower processed demand and wallet counts. The sources reviewed do not establish how the participation measures reconcile, so the two sets of figures should not be collapsed into one measure of participation.

A month later, a different campaign exposed a different problem. The USDm pre-deposit program ran into configuration, KYC and transaction-control failures and was abandoned; MegaETH announced that all roughly $500 million would be returned once a refund contract passed audit. It concerned stablecoin deposits, not the October MEGA auction. Fast execution had not eliminated the ordinary operational risks of launching financial products around the chain.

Public mainnet opened on February 9, 2026 with more than 50 applications reported live. The network used ETH for gas, one active sequencer for ordering, EigenDA for data, and Ethereum for settlement. MEGA still did not exist as a transferable mainnet token.

The team had promised to wait until one of four performance milestones was achieved. On April 23, ten Mega Mafia applications met the first threshold, including a requirement for meaningful on-chain interactions. That event started a seven-day clock, and MEGA launched on April 30. The launch rule turned an accelerator portfolio into evidence that the network was ready.

The same idea shaped supply. Detailed disclosures reserve 53.3% of the ten-billion-token design for KPI rewards. Holders can commit MEGA in Flux, and future tranches are distributed when milestones for reliability, performance, adoption or decentralization are attested. In the current phase, the Foundation defines and verifies those targets.

MEGA’s other advertised roles do not all exist yet. ETH remains the gas asset. Proximity Markets, sequencer staking and rotation, MEGA gas payments and DAO governance are described as future layers. USDm creates a nearer economic link: income the Foundation receives from the stablecoin is used to buy and accumulate MEGA, rather than paid directly to holders.

The token contract itself took a separate path. It launched with an owner-only mint function and no coded hard cap, then renounced ownership on June 19. At the September snapshot, the owner and proxy slots were zero and the supply was just under ten billion after burns. That limits token-contract intervention, while the rollup’s core contracts remain under a different upgrade regime.

On July 16, less than three months after TGE, MegaETH said there would be no third Mega Mafia cohort. Kong wrote that the program had helped about twenty teams raise roughly $80 million but that most successful applications were no longer building with MegaETH. The team had taken no equity or governance stakes in those companies because it expected voluntary alignment; little of the created value returned to the network.

MegaETH redirected resources toward first-party products and “OMEGA” applications designed around its own wallet, stablecoin and latency. The change does not erase the fact that ten apps were live in April. It changes what that fact can prove. Milliseconds and processed transactions measure a machine over an interval. A deployed-app count can trigger a token, but it cannot guarantee that builders, users or economic value will stay.