Starknet’s L1 research poll closes at 93.5% Yes, with the Foundation-delegated list casting about 94% of the power

Update, 22 October 2026, 11:30 UTC: the poll closed at 09:32 UTC today. Final Snapshot figures give 93.52% Yes, 4.57% No and 1.91% Neutral, across 3,236 votes and 489.3 million STRK of voting power. The Foundation-delegated whitelist still supplied about 94% of that power, and most of the No weight came from two of its addresses, which voted on 21 October. StarkWare chief executive Eli Ben-Sasson says a draft whitepaper should be out for public review in November, and any move to a layer 1 would still need a separate vote. The article below has been revised; figures dated 17 October are the mid-vote snapshot taken when it was first published.

Starknet’s community has backed research into leaving Ethereum to become an independent, post-quantum layer 1. When the poll closed at 09:32 UTC on 22 October, the option “Yes, this should be researched” held 457.6 million STRK of voting power, or 93.52% of the total, according to the proposal’s Snapshot page. No took 4.57% and Neutral 1.91%, across 3,236 votes. The poll opened on 15 October at 09:32 UTC, had no quorum and is non-binding: it asked only whether research should go ahead.

The base behind that majority is narrow. Coindaro’s count of the final Snapshot vote data shows that about 94% of the voting power cast, 461.1 million STRK, came from a whitelist labelled “STRK (SNF delegated)”: voting power the Starknet Foundation has assigned to 109 addresses with fixed weights. 55 of those addresses voted. Holders voting with their own staked or held STRK supplied most of the rest.

The Yes share was 99.95% at 07:09 UTC on 17 October, when this article was first published. It fell to 93.52% mainly because two whitelisted addresses voted No on 21 October, the day before the close, with 16.67 million and 5.56 million STRK.

The question of Starknet quantum resistance moved fast: a post on X on 8 October, a formal poll a week later, a result on 22 October. STRK, the network’s token, rose more than 80% into 11 October and has since given those gains back. It traded at $0.0534 at the close, 09:32 UTC, and at $0.0550 at 10:45 UTC on 22 October, down 2.2% in 24 hours, according to DefiLlama price data. There was no visible jump at the result.

Starknet quantum resistance poll: a vote on research, not on leaving

The proposal, titled “Should Starknet become an independent post-quantum L1?”, sits in the starknet.eth space on Snapshot, a widely used off-chain voting tool. Voters pick one of three answers: “Yes, this should be researched”, “No, this shouldn’t be researched” or “Neutral, I need more information”. There is no quorum. Voting power was fixed at Ethereum block 26,196,434.

The text is careful about what a win would mean. It describes itself as a “non-binding poll” meant “to gauge STRK holders’ interest in exploring that direction further.” A Yes vote, it says, “indicates support for further research. It does not constitute approval of a transition to an L1.” Any actual move “would be subject to a dedicated governance process.”

The forum post that accompanies the poll, published on 15 October by henrli of the Starknet Foundation, sets out the reasoning. The parts of Starknet the project controls itself can become quantum-resistant soon. Two pieces cannot, because they depend on Ethereum’s own timeline: the bridge and data availability. An independent layer 1, the post argues, would remove that dependency.

Some things would not change, according to the same post. STRK would remain the network’s token, with no new token planned, and the bridges from Ethereum to Starknet would stay open. The post also maps out the three outcomes. A Yes leads to research, then more detailed proposals and further polls. A No means post-quantum work carries on with Starknet as a layer 2. Neutral keeps the discussion open. When the poll opened, the post said no timeline existed yet.

The author’s summary of where Starknet stands today is short: “The core protocol is provably safe from quantum computers.” Readers new to the vocabulary can start with our explainer on what a layer 2 network is, the category Starknet belongs to today.

Who cast the 93.5%: the Foundation-delegated whitelist carried about 94% of the power

Snapshot counts voting power from four sources in this poll: vSTRK on Starknet, STRK held on Ethereum, staked STRK, and the whitelist labelled “STRK (SNF delegated)”. That whitelist holds 109 addresses, each given a fixed weight of 16.67 million, 9.26 million or 5.56 million STRK. Together they carry 931.5 million STRK of potential voting power.

Snapshot’s final tally at the close, 09:32 UTC on 22 October, next to the mid-vote figures this article reported on 17 October:

Snapshot records, for each vote, how much power came from each source. Coindaro added them up across all 3,236 votes, 489.3 million STRK in total.

The delegated share barely moved over the week, even as the number of voters doubled. The 10 largest voters cast 28.1% of the power, down from 39.8% on 17 October. Each of the six largest votes carries about 16.67 million STRK from an address on the whitelist: five voted Yes, one voted No. In all, the whitelist used 461.1 million of its 931.5 million STRK, about half of its weight, and 52 of its 55 voting addresses chose Yes.

The No side grew from four wallets on 17 October to 19. Two whitelisted addresses account for 22.2 million of its 22.4 million STRK; the remaining No votes are small, the largest holding about 133,010 and 22,000 STRK. Six voters chose Neutral, and one whitelisted address with 9.26 million STRK makes up almost all of that option’s 9.33 million.

By head count, turnout was the highest Starknet has recorded. The project’s survey polls of March and April 2026, which closed on 6 April, drew roughly 343 to 350 votes each and about 490 to 500 million STRK of voting power. This poll drew 3,236 votes, more than nine times as many, for a similar total of voting power. StarkWare called it the “highest participation in Starknet’s governance history”, Cointelegraph reported. The Block, citing StarkWare, put participation at more than 2,800 wallets, a lower figure than the vote count on Snapshot.

Yes led from the first hours. The Block counted 177.6 million STRK at 99.91% Yes at 14:26 UTC on 15 October, five hours after the open.

Two things are not visible on the Snapshot page. The first is the current rule behind the delegate weights. In March 2025, the Foundation set delegated voting power at 35 million, 10 million or 4 million STRK per delegate across 180 delegates and 1.7 billion STRK. Today’s whitelist uses different tiers, 16.67 million, 9.26 million and 5.56 million, across 109 addresses, and Coindaro found no published post explaining the change. The second is how individual delegates reached their decision on this poll: nothing published shows it, and nothing suggests they received voting guidance.

The delegation scheme itself is not new. When the Foundation set up its three tiers in March 2025, it gave the voting power to community delegates, including former members of its Builders Council, and said the aim was for decisions to be led by major contributors to the ecosystem rather than by the Foundation. Delegates who stop taking part can lose their allocation.

From a Token2049 post to a Snapshot poll in seven days

The idea surfaced on 8 October. Eli Ben-Sasson, co-founder and chief executive of StarkWare, the company that built Starknet, asked on X whether “Starknet becoming an L1 for post-quantum agility” was a good idea, tying the post to his Token2049 talk that day. He wrote that the quantum threat “may be MUCH closer than we think” and called for “bunker mode”. Minutes later, as Unchained reported, Starknet’s own account said it was “actively considering becoming an L1” as a path to “the first fully quantum-resistant network”, with a 2027 target.

When the poll opened, that 2027 date appeared only in those posts, and the forum proposal of 15 October said no timeline existed yet. After the close, Ben-Sasson restated 2027 as StarkWare’s target for a quantum-safe Starknet (see the last section).

The poll followed a week later. Ben-Sasson told Cointelegraph on 15 October: “Quantum-proofing crypto is urgent, so we must move at lightning speed.” StarkWare told The Block the same day that “the parts of Starknet under Starknet’s control can be quantum-resistant soon”, adding: “No timeline for that exists yet, but Starknet starts further ahead than most chains”.

On 16 October at 17:33 UTC, The Defiant published a podcast with Ben-Sasson titled “STRK’s L1 Bet: Eli Ben-Sasson on Starknet’s Exit Plan”. Its page lists topics but no answers, so it adds nothing on the record yet.

The June roadmap did not mention an L1

StarkWare’s post-quantum plan predates the poll. On 30 June 2026, the company published a three-phase roadmap, reported by The Block. Phase 1 replaces Pedersen hashing with BLAKE2 for state commitments, contract addresses and the configuration hash, and adds post-quantum consensus signatures such as Falcon-512. Phase 2 builds tools to migrate older contracts. Phase 3 deals with what depends on Ethereum: the bridge messaging syscalls, which use secp256k1 and r1 elliptic curves, and the KZG commitments used in EIP-4844 blobs.

Ben-Sasson said then that the plan was “our path to making Starknet a safe haven for funds whatever quantum may bring.” That roadmap did not raise leaving Ethereum. All protocol-level changes still need governance approval, Unchained noted on 8 October.

Is Starknet quantum resistant? The proofs are, the Ethereum link is not

A layer 2 borrows three things from Ethereum. It settles there, meaning Ethereum verifies its proofs and holds the final record. It posts its data there, so anyone can rebuild its state; Starknet has sent state differences to Ethereum as blobs since 13 March 2024, according to L2BEAT. And its bridge lives there: assets moved onto Starknet are locked in Ethereum contracts.

Quantum computers threaten these pieces unevenly. Shor’s algorithm, run on a large enough quantum machine, would break elliptic-curve cryptography. That covers the ECDSA signatures on most wallets and the elliptic curves inside most SNARK-based rollups. Grover’s algorithm only speeds up brute-force searches, and slightly larger hash parameters offset it, StarkWare explains in a 20 July blog post.

Starknet’s proofs are STARKs, a type of zero-knowledge proof. “STARKs rely only on post-quantum secure hash functions for their cryptographic commitments,” StarkWare writes. They also need no trusted setup. Since 19 October 2025, Starknet has generated them with the Stwo prover, L2BEAT records.

Accounts are the second piece Starknet can change on its own. “Every Starknet account is natively a blockchain that runs exactly as written when it is called, holding and moving funds according to its code, without an intermediary able to change the outcome.">smart contract, and account abstraction is the only account model,” the StarkWare blog says. An account can therefore accept a post-quantum signature without a network fork. Falcon-512 accounts already exist, one from S2morrow written in Cairo and one from OpenZeppelin, though the blog warns that “The OpenZeppelin contracts are early implementations and haven’t been audited yet”. Post-quantum signatures also cost more per transaction.

What Starknet cannot fix alone is everything anchored on Ethereum. The bridge messaging syscalls use elliptic curves, and the blobs that carry Starknet’s data rely on KZG commitments, which use elliptic-curve pairings. The StarkWare blog says it plainly: “Starknet’s full-stack quantum safety depends on Ethereum completing its own transition.”

Ethereum has a plan, on a longer clock. The ethereum.org quantum-resistance page, updated on 8 September 2026, targets core post-quantum infrastructure “by approximately 2029” and calls these “planning milestones, not guaranteed commitments”. Ethereum intends to replace BLS signatures in consensus with leanXMSS, a hash-based scheme, and KZG with STARK- or lattice-based commitments. Hash-based signatures are much larger: about 3,000 bytes against 96 for BLS. The Ethereum Foundation set up a dedicated post-quantum team in early 2026. The same page tells readers: “Your funds are safe today.”

Starknet says it can move on its own pieces now. Ethereum’s date for the rest is around 2029, and that gap is the argument the Foundation puts to voters.

What leaving Ethereum would cost, according to the forum

The forum thread is far less one-sided than the vote. A validator posting as Cumulo backed the idea and offered to run test nodes. A user named ceri1 replied on 16 October with the sharpest objection: “this is a transfer of dependency, not a gain in decentralization”. In ceri1’s view, an independent chain would lose the security budget Ethereum provides, and the proposal leaves open questions on finality, economic security, bridge risk and migration.

Another participant, boomelant, asked whether STRK staking could stand in for Ethereum’s security. henrli called these “valid questions” and said research could help answer them if the vote is positive.

None of those answers had been published when the poll closed on 22 October; StarkWare says a whitepaper draft is due in November. Nothing has been published on how a Starknet L1 would reach consensus, who would run its validators, how fast it would finalise blocks, how large its security budget would be, or how a bridge back to Ethereum would work.

For comparison, L2BEAT classes Starknet as a Stage 1 ZK rollup. A Security Council of 12 members can push emergency upgrades instantly with 9 signatures, while normal upgrades carry an 8-day delay. Total value secured stood at $585.8 million on 17 October, about half of it in STRK. Starknet’s DeFi total value locked was $160.2 million on 17 October, according to DefiLlama, against $167.1 million on 11 October and a peak of $330.8 million on 26 March 2024.

Layer 2 networks are rethinking their ties to Ethereum

Starknet’s poll lands in a year of doubts about the layer 2 model. Vitalik Buterin wrote in February that progress towards stage 2 had been “far slower and more difficult than originally expected”, Cryptopolitan recalled on 15 October. ENS dropped its planned Namechain rollup and stayed on Ethereum mainnet. Abstract will shut down on 15 December, and Blast is winding down. Most of those moves went back towards Ethereum. Starknet, Cryptopolitan noted, would be the first major L2 to explore leaving in the other direction.

Quantum risk has also climbed up the industry’s agenda this week. We covered Google’s latest quantum error-correction result on 16 October. Separately, the Quantum Alliance, an industry group with nine members, launched in Lisbon on 15 October; StarkWare is not among them. No direct reaction from Ethereum researchers or the Ethereum Foundation to Starknet’s poll had been reported by 17 October.

Starknet is also losing its largest app. The perpetuals exchange Extended moves its settlement from Starknet to Circle’s Arc chain at 13:00 UTC on 22 October, as we reported in our piece on Extended’s move to Arc.

STRK trades about 22% below its 17 October price

STRK traded at $0.0550 at 10:45 UTC on 22 October, down 2.2% in 24 hours, according to DefiLlama. It stood at $0.0534 when the poll closed at 09:32 UTC. That is about 22% below the $0.07074 recorded at 07:00 UTC on 17 October, when this article first appeared. Coindaro has found no cause to attach to the move.

On 17 October, STRK was down 14.81% in 24 hours and up 0.32% over seven days. CoinGecko showed a 24-hour range of $0.06973 to $0.08305, a market capitalisation of $537 million (rank 111) and $154 million of daily volume at about 07:10 UTC that day.

The token had rallied hard after the 8 October posts. It rose more than 80% in a week to about $0.126 on 11 October, its highest level since December 2025, according to Cryptopolitan. On the day the poll opened, CryptoTimes put it at $0.08867 at 17:15 UTC, up 15.3%, with $401 million of daily volume.

Coindaro found no confirmed reason for that 17 October fall. There was no token unlock that day. The Starknet status page shows no incident between 7 and 17 October. No Binance delisting or monitoring tag was found, and there was no negative news from the project. A technical note from DiarioBitcoin published on 17 October also describes the catalyst as unconfirmed.

What the data shows is a broader pullback among that week’s winners. WAL lost 15.76%, ZRO 13.45%, PYTH 10.94% and ZK 10.85% over the same 24 hours, while BTC and ETH slipped 0.74% and 0.77%. Total crypto market capitalisation stood at $2.76 trillion, down 3.48%. That pattern is an observation, not an explanation.

Supply is the other background factor. On 15 October, 127 million STRK reached early contributors and investors in the monthly token unlock, 1.27% of the 10 billion maximum supply, according to CryptoTicker. Circulating supply stood at 7.591 billion STRK on CoinGecko on 17 October. The monthly schedule runs until 15 March 2027.

Nothing in this article is investment advice.

What comes next: a whitepaper draft in November, then new votes

Nothing changed on the network when the poll closed. Starknet remains an Ethereum layer 2, and any actual move to a layer 1 would need a separate, binding vote, according to The Block and Cointelegraph. No date has been agreed for such a transition. StarkWare itself did not vote, The Block reported.

The research phase has started. Ben-Sasson said StarkWare hopes “to release a draft for public scrutiny in November”, meaning a whitepaper on the transition. He also set 2027 as the target for Starknet to become quantum-safe, arguing that “Ethereum’s target of 2029 to become quantum-safe is too late, and deadlines slip”. The forum post had already set out the sequence for a Yes: research, then more detailed proposals and new polls.

Some points are fixed whatever the outcome, the Foundation said: STRK stays the network’s token, no new token will be issued, and the bridges from Ethereum to Starknet stay open. The open questions from the forum thread, from validators to the bridge back to Ethereum, are the ones the November draft will have to answer.

The next STRK unlock falls on 15 November: 127 million tokens, 66.61 million to early contributors and 60.39 million to investors, according to CoinGecko.