Testnet Milestones: First Post-Quantum Payments and 2-of-2 Multi-Sig Confirmed On-Chain

Blythex

September 18 was a big day for AntiQua.

In less than two hours, our public testnet crossed three milestones – every transaction signed with ML-DSA-87 (Dilithium), the NIST post-quantum signature standard, mined into real blocks and accepted by our public archive node.

TL;DR

First post-quantum payment confirmed on-chain
First 2-of-2 multi-signature wallet funded
First multi-sig spend – two independent post-quantum signatures, verified and confirmed

First single-signature spend

18 Sep 2026, 16:06 CEST

A simple payment of 10 AQA from a mining payout address to a fresh AIP-39 wallet (our 24/36/48-word recovery standard), with a 0.01 AQA tip and a 0.001 AQA relay fee.

Path: wallet send → signed transaction → mempool → mined → confirmed by the archive node
Transaction size: ~9.4 KB (post-quantum signatures are big – more on that below)
Confirmed in block #112
Receiver balance after confirmation: 10.000000 AQA

tx     9e353fb5ba5c1af53ec726946a4ef41a93c3152bc652a89148b8281f4388600a
block  #112  ea98a8b5bf5ca0288d277f96300bd841ad5fd06ada2c51c4e41cef8ee947ef9b

Multi-sig wallet funded

18 Sep 2026, 16:46 CEST

We created a 2-of-2 multi-signature wallet on testnet – funds can only move when both key holders sign – and sent 10 AQA into it.

tx     2d4df3e35cabbd89ef3da8b4fb4f56a31f375e1761a34ad169f464156f88d60e
block  #127  831935a282b0a417baedefd68b097ad5a9aade33a64e619e047a7ddb3f3f3320

First multi-sig spend

18 Sep 2026, 18:04 CEST – the big one.

Two signers each loaded their own wallet, built one transaction together and signed it independently with their own ML-DSA-87 keys. The node checked both signatures before accepting it:

verify_multisig_input [SUCCESS]: enough signatures present (2/2 required)

The spend went into the mempool, was mined into block #130, broadcast to the network and accepted by the archive node.

tx     d57ab29312337aa64958f84fefcfc6141487e666eb09cfcb0529706e08d1544f
block  #130  ccb98b6b6438dff6e76492d624e777d454c7d18262e136109638bffc242180a7

And the numbers add up to the last decimal:

Wallet	Balance	Math
Receiver	15.000000 AQA	10 (milestone 1) + 5
Multi-sig	4.989000 AQA	10 − 5 − 0.01 tip − 0.001 fee

For the developers among you, this is the whole flow in our node CLI:

walletload <signer-1>
walletload <signer-2>
mscreate 2 testnet
mstx ms 2 <receiver> 5 0.01 0.001
mssign last 0 0
mssign last 0 1
mine 1

What we had to fix to get there

Multi-sig didn't work on the first try – the archive node rejected our transactions with a hash mismatch. We tracked it down to two independent bugs:

Fee set after hashing. The multi-sig builder calculated the transaction hash first and set the fee afterwards – so the hash no longer matched the transaction body. Now the fee is set first, then the hash is computed.
Multi-sig fields lost on the wire. The network serializer only sent the basic input fields and silently dropped the multi-sig data (required signatures, signer keys and signatures). The receiving node filled in defaults and computed a different hash. You can see it in the numbers: the transaction packet grew from 656 bytes (broken) to 15,232 bytes (complete), and the block from 12,888 to 27,464 bytes.

The fix: hashing, the mempool and the network now use one single serialization path, so they can never disagree again. A dedicated regression test makes sure the truncated format is rejected and the full format round-trips cleanly.

Why these sizes matter

A post-quantum signature is much larger than a classic Bitcoin signature – that's the price of security against quantum computers. One single-sig payment is ~9.4 KB, a 2-of-2 multi-sig transaction ~15 KB.

This is exactly why AntiQua was designed from day one with compact commits and checkpoints to keep the chain lean over time. If you missed it, read Why Post-Quantum? The Threat is Real.

What this does not prove yet

We'd rather be honest than hype. These milestones were done with the node CLI on testnet. Still ahead:

Multi-sig in the desktop wallet (GUI)
Multi-sig in the Android light wallet against the live chain
Larger setups like 2-of-3 or 3-of-5
Compact block relay (currently switched off)
Mainnet – of course

We'll keep posting progress in the Core Changelog.

Questions about the transactions, multi-sig or post-quantum signatures? Ask below – we're happy to go into the details.

— Blythex