Introducing Strongpoint MPC: fast threshold signing for hash-based post-quantum signatures
Strongpoint MPC is Project Eleven's first post-quantum MPC product: fast, dealerless threshold signing for hash-based signatures, starting with Ethereum's leanSPHINCS and leanXMSS.
Contents
Introduction
Today we are announcing Strongpoint MPC, the first post-quantum MPC product from Project Eleven. It lives inside Strongpoint, our existing custody product, and it is the start of a broader suite. Over time the suite will support every post-quantum signature scheme our customers may need. We are starting with the one that many people considered the hardest to get right in a threshold setting: hash-based signatures.
Where the conversation started
When the industry first began talking about the post-quantum transition for digital assets, most of the debate was about which family of cryptography to pick. Hash-based signatures, lattice-based signatures and more exotic constructions such as isogenies were all on the table.
Hash-based signatures came with a reputation but people saw three main problems. The signatures are large. They are awkward to use with hardware signers. And they were thought to be very hard to use in threshold schemes such as MPC. Earlier work estimated that the secure hash evaluations alone for a threshold SPHINCS+ signature would take approximately 85 minutes using contemporary garbled-circuit techniques, before accounting for the rest of the signing protocol.
Why that looked like a dead end
An hour per signature makes a product impossible. Those early estimates were based on SPHINCS+ parameters as later standardized by NIST in FIPS 205 as SLH-DSA. Those parameters are already too large for most blockchain use but even with smaller, more blockchain friendly parameters, the MPC numbers were still poor.
This is where our research and development started. We wanted to understand how to build a hash-based post-quantum MPC product that could actually serve teams adopting SPHINCS-based signatures and constructions inspired by SPHINCS+. That work led to a series of breakthroughs in threshold signing for hash-based schemes, and those breakthroughs are what make Strongpoint MPC possible.
What Strongpoint MPC can do
Strongpoint MPC brings our research into a working threshold signing system. The initial implementation demonstrates three capabilities: fast signing for SPHINCS-inspired constructions, support for common threshold configurations, and dealerless key generation.
We can produce SPHINCS-inspired signatures, like the ones expected for account transaction signatures on Ethereum, in a fraction of the time earlier work suggested. Depending on the setup, a full threshold signature takes less than a second. That is a large step beyond any published threshold result for SPHINCS-style schemes, and it comes from our own variant, SPHINCS-G, which we will describe in a dedicated paper soon.
Fast signing only matters if it fits the way teams split control over their keys. Strongpoint MPC already covers the threshold setups teams normally use in practice, with the usual numbers of participants. Another part of our R&D, already underway, will extend this to arbitrary thresholds.
Strongpoint MPC is also dealerless. No single party ever generates or holds the full key, even during setup, so there is no trusted dealer to rely on or to attack. For teams that do want a dealer, the same stack can support that setup too.
We will share them soon in dedicated papers, showing how they apply to MPC and to other areas that matter for digital assets.
Why we are starting with Ethereum
Strongpoint MPC is built for digital assets in general. We are starting with Ethereum because it is currently the most advanced major network in its hash-based post-quantum roadmap, so our support begins by following what Ethereum is building.
By support we mean the real signature, the same one the network will adopt. We do not add a custom verifier or custom systems that reduce user privacy. A transaction signed with Strongpoint MPC looks like any other transaction signed with the network's own post-quantum scheme.
Why hash-based signatures, and why now
Hash-based signatures are one of the most interesting kinds of post-quantum cryptography. They also rest on some of the most conservative security assumptions available. That makes them a strong foundation for the future of Ethereum.
They are simple, and the old problem of signature size is now being addressed at the network level. Recent proposals for signature aggregation using STARKs and dedicated proofs of knowledge shrink the on-chain cost of large signatures. Research progress has made these signatures viable for the network.
The clearest example is EIP-8288 (opens in a new tab), proposed by Vitalik Buterin. It aggregates signatures and STARK proofs recursively in the mempool, so a block carries one composite proof instead of every signature. The proposal estimates that a hash-based signature verified directly costs two to three kilobytes and 150,000 to 200,000 gas. Aggregation moves much of this cost into a shared proving process, reducing the per-signature cost borne on-chain. On the consensus side, EIP-8292 (opens in a new tab) covers aggregators for hash-based validator signatures.
Aggregation takes care of the network. What remained open was how institutions hold these keys and sign with them. We make these signatures usable for institutional custody, and for sophisticated custody setups more broadly, through MPC. Strongpoint MPC supports both leanSPHINCS and leanXMSS, so teams get threshold security on either scheme without giving up signing speed.
Offering post-quantum MPC security for every scheme
Strongpoint MPC is designed as a suite. Our hash manipulation techniques and our cryptographic and engineering results carry over to other hash-based signature schemes, as long as they remain compatible with the stack we are building.
At Project Eleven we see the post-quantum and post-AI transition as an opportunity. Public blockchains can come out of it more secure and more future proof than they have ever been. Getting there takes more than new cryptography. Institutions need custody that works well at every step, without stitching together a different tool for each new scheme.
That is why we are building Strongpoint as a single, unified product. Our goal is to offer the best institutional custody experience at all times, through every stage of the transition and with whatever signatures the networks adopt along the way.