MegaETH shuts its Mega Mafia accelerator, admitting most of its hits build elsewhere
MegaETH is ending its Mega Mafia accelerator after two cohorts, with team member Shuyao Kong conceding that most of the roughly 20 startups it incubated have left for rival chains.
Ethereum scaling project MegaETH is winding down Mega Mafia, the flagship accelerator it ran for two years across two cohorts, according to a Thursday X post from core team member Shuyao Kong reported by The Block. The stated reason is blunt for a program the company once put at the center of its ecosystem: it worked for the startups, not for MegaETH.
The program hand-picked roughly 20 teams to "co-live and co-build" with MegaETH's core team, which supplied technical, managerial and market-making help. Those teams went on to raise about $80 million in pre-seed to Series A funding, per Kong. Notably, MegaETH took no equity, ownership or governance position in any of them — a choice Kong framed as fostering "genuine alignment," and which now leaves the company with little to show for the incubation.
Kong's own summary: "In many ways, the Mega Mafia was the best incubator of this cycle. But very little of that value has trickled to Mega," adding that of the successfully incubated projects, "most of those applications are no longer being built with us." The Block's reporting bears that out. Global Token Exchange (GTE), a flagship Mafia project that has raised $25 million across rounds, said it had "grown up" and would build its own chain. Noise, a social "attention market," launched on Base after a $7.1 million seed led by Paradigm in January 2026. HelloTrade went to Monad. Two MegaETH-backed projects, Avon and Valhalla, appear to have shut down. Cap, a stablecoin project that launched on MegaETH in early 2026, is pursuing a multichain strategy.
Instead of a Mega Mafia 3.0 cohort, Kong said MegaETH will redirect funding to "first-party applications" — consumer apps it builds and owns itself, keeping "all the upside and accountability."
Key facts
- Mega Mafia ran two years / two cohorts; ~20 teams incubated (The Block, citing Kong's X post, 2026-07-16).
- Portfolio raised ~$80M pre-seed to Series A; MegaETH took no equity or governance (Kong via The Block).
- Departures: GTE (own chain, $25M raised), Noise (Base, $7.1M seed led by Paradigm, Jan 2026), HelloTrade (Monad); Avon and Valhalla appear defunct; Cap going multichain (The Block).
- MegaETH mainnet launched February 2026; MEGA token launched April 2026 after 10 apps went live (The Block).
The real-world read
This is a public admission that the incubator's economics didn't work for its sponsor — and the honesty is worth crediting. But note who's framing it: the concession that "most" successful apps left comes from MegaETH's own team member, in a post pitching the pivot to first-party apps, so the walk-away figures are the interested party's own characterization, not independently tallied. The decision to take no equity, once sold as principled alignment, is now retroactively cast as the reason "little value trickled" back — the same choice, opposite spin. And "most" does soft work: The Block names concrete exits (GTE, Noise, HelloTrade) but the full retention count isn't disclosed. Whether "first-party apps" fares better is an assertion about the future, not a result.
This is news coverage, not financial advice.
Sources
- The Block (Daniel Kuhn), 2026-07-16 — the sole report, quoting Shuyao Kong's X post and detailing project departures, funding figures, and MegaETH's mainnet/token timeline. Independent secondary outlet; discloses Foresight Ventures as majority investor. Kong's X post is the primary source it draws on.