Ethereum co-founder Vitalik Buterin wants to move complex cryptographic operations off the blockchain entirely and onto users’ own devices — a draft proposal called EIP-8288 that could shrink what every node has to process, if it ever ships.
By Amir Hassan | October 19, 2026
On the social platform X, Buterin called for a movement toward “EIP-8288 maximalism,” arguing that individual devices should handle the heavy cryptographic lifting so Ethereum never has to add another built-in cryptographic function again. The idea, co-authored with Thomas Coratger and dated June 3, 2026, was introduced at the ETHTaipei 2026 event under the Ethereum Foundation banner. It remains a draft — but it points at where the network’s leadership wants to go.
The Hook: Ethereum Has a Bloat Problem
Every time Ethereum wants to support a new kind of cryptography, it historically adds what is called a precompile — a built-in function that every node on the network must maintain forever. Think of a precompile like a permanent appliance installed in every kitchen in the country: convenient for whoever needs it, but every new one makes all the kitchens heavier and harder to move. Each addition grows the chain’s overall size and permanently complicates the software every node runs.
EIP-8288’s answer is radical subtraction: stop adding precompiles altogether, and let users’ own phones and laptops do the complex math instead.
How the Lean Plan Actually Works
The proposal centers on STARK proofs — concise mathematical constructs that verify a calculation’s accuracy, letting nodes trust the outcome without repeating the computation. Under the draft, each user’s device would generate a STARK proof locally. Nodes in the mempool — the waiting room where transactions sit before inclusion — would then aggregate and bundle these proofs, so that each block carries only a single proof that validates all of them.
In plain terms: instead of the network checking everyone’s homework line by line, every student checks their own work at home and hands in a tamper-evident summary card, and the teacher only reads one combined card per class. The block stays small, the verification still holds mathematically.
The Numbers: From 128 kB Monsters to 3,000 Gas Checks
The appeal is the size disparity. According to the draft as reported by 24/7 Wall St., current quantum-resistant signatures require roughly 2 to 3 kB of data and cost 150,000 to 200,000 gas to verify on-chain, while full STARK proofs demand over 128 kB. Gas is the unit Ethereum uses to measure computational work, and users pay for it in ETH.
- 150,000–200,000 gas — current on-chain verification cost for quantum-resistant signatures
- Over 128 kB — the data weight of a full STARK proof stored on-chain today
- 3,000 gas — proposed verification cost for leanSPHINCS signatures under EIP-8288
- 30,000 gas — proposed verification cost for leanSTARK proofs
That is a drop of roughly an order of magnitude or more for the network’s verification load — the difference between weighing every suitcase at check-in versus trusting a sealed security scan each traveler already did at home.
The Quantum Subtext
The push is not happening in a vacuum. The cryptographic schemes in question — like SPHINCS-based signatures — are the quantum-resistant kind, built on mathematical problems that future quantum computers are expected to struggle with. But quantum-resistant cryptography is notoriously heavy. The old approach would have meant stuffing big new precompiles into Ethereum to make it quantum-safe; EIP-8288’s bet is that user devices are now powerful enough to carry that weight instead of the chain. It is a design choice with a deadline attached, because the quantum clock keeps ticking for every blockchain.
The Catch: It Is a Draft With Holes In It
Before anyone marks a calendar, the caveats are substantial. Two core parameters, AGGREGATED_VK and DEPS_HASH, are literally marked “TBD” in the specification. The proposal is not yet tied to any specific network upgrade, no client team has committed to building it, and no wallet or privacy application has announced plans to adopt it. It also depends on EIP-8141, a separate proposal that lays its groundwork. Ethereum’s recent history is full of promising drafts that waited years for a network slot — or never got one.
What This Means for You
For ETH holders, the stakes are indirect but real. Leaner verification means lower fees and a lighter chain — both good for long-term network health. The trade-off flagged by analysts is subtler: a portion of every transaction fee is permanently burned, removing ETH from circulation, so drastically cheaper verification could also mean less ETH burned per activity. The draft itself provides no estimates for how that balance nets out. If you use Ethereum wallets, a future like this could eventually mean your device quietly does more cryptographic work in the background — slightly more computation on your phone, much less bloat on the chain.
The Verdict
EIP-8288 is a direction, not a delivery. But it signals that Ethereum’s leadership wants the network to stay lean as quantum-resistant cryptography — unavoidably heavy stuff — arrives over the next few years. The technical idea of users proving things locally while blocks only carry aggregated proof is one of the cleaner escapes from the precompile trap anyone has proposed. Whether it graduates from draft to mainnet will depend on client teams, wallet makers, and the slow machinery of Ethereum governance.
The cryptocurrency market remains highly volatile. This article is for informational purposes only and does not constitute financial advice.
generating STARK proofs on a phone battery before every transfer sounds great until your device is 4 years old and the proof takes 40 seconds. whats the fallback
@zk_grannie honestly the 40 second proof thing is overblown, bloom filters already do most of it client side. the heavy STARK stuff would only hit weird edge cases, not every transfer
fair question but the status quo is worse. every precompile is a permanent tax on every node forever, 128kB monsters included. something had to give
eip-8288 maximalism is the most vitalik phrase ever coined. man really said lets fix chain bloat by deleting the chain
june draft, ETHTaipei presentation, vitalik pushing it on X. ive seen this movie before, it lands as a watered down version in some 2028 fork and everyone claims victory
Moving verification to user devices makes sense until you remember half the phones in circulation have broken secure enclaves or ancient android versions. Whose device fails first gets yelled at on Twitter.
@Ville the device attestation layer is supposed to cover broken enclaves, and honestly a degraded four year old phone verifying its own proofs still beats every node on earth chewing 128kB blobs forever
@Elif degraded phones beating 128kB blobs forever only works while verification actually runs. one buggy OEM update silently skipping the check and nobody notices for weeks. the attestation story gets messy the second someone ships a spoofer
the precompile analogy is perfect honestly. every new one is furniture bolted into every kitchen forever. no wonder node operators are cheering this
Calling it EIP-8288 maximalism is doing a lot of marketing work for what is basically a draft from June. Half of these proposals die in comments before Devcon.
@Tarek sure, but this one came out of ETHTaipei with Vitalik and Coratger both on it. That pairing usually survives to at least a testnet
@Tarek a june draft that vitalik demos at ETHTaipei with coratger co-authoring does not usually die in review. the anza crowd is clearly at least listening this time
Remember when every Ethereum upgrade meant two years of arguing about adding yet another precompile. If this draft actually ships, those fights end. Slowly, but they end.
and but the difference is these fights end by deleting code instead of adding it. thats the whole point of the maximalism framing, fewer permanent protocol taxes