Tezos Quantumnet is an experimental test network where Tezos developers run post-quantum replacements for the blockchain’s cryptography side by side: ML-DSA-44 account keys, hash-based consensus signatures, a new way to assign baking rights and a post-quantum data availability layer. It went live in September 2026, announced on Tezos Spotlight on September 24. It is not a mainnet upgrade, and nothing on it has value.
Last checked: November 10, 2026. On that day Quantumnet’s public RPC was still producing blocks (head at level 727,301 at 05:27 UTC), and its network configuration still named the chain it launched with, so the network is in its first iteration. We found no Tezos protocol proposal that would bring these components to mainnet.
What Quantumnet replaces
Tezos uses cryptography in more than one place, and each place needs its own replacement. The table compares Tezos mainnet today with the first Quantumnet iteration, based on the Spotlight announcement and the Nomadic Labs launch post on Tezos Agora.
| Part of the protocol | Tezos mainnet today | Quantumnet, iteration 1 |
|---|---|---|
| User (manager) keys | Elliptic-curve signatures; tz5 exists only behind a feature flag | tz5 accounts with ML-DSA-44 (NIST FIPS 204) |
| Baker consensus signatures | BLS signatures, which can be combined | tz6 keys with XMSS, a stateful hash-based signature |
| Attestation aggregation | Combined BLS signatures to save block space | Post-quantum STARK aggregation from the LeanEthereum project |
| Baking rights | Randomness-based lottery with a Verifiable Delay Function | Smooth Weighted Round Robin, no protocol randomness |
| Data Availability Layer | KZG commitments and proofs | ZODA, built on hashes and erasure coding |
| Sapling and timelocks | Available in Michelson | Left out of this iteration |
The tz5 account type came first. Tezos’ Ushuaia upgrade, activated on mainnet on June 30, 2026, added tz5 support but kept it switched off on mainnet so wallets and custodians could prepare. Quantumnet is the first testnet to switch it on.
Why bakers need two kinds of keys
The tz5/tz6 split is the clearest sign that this is more than a signature swap. Tezos combines many consensus attestations into one to save space, and ML-DSA does not support that kind of aggregation. So Quantumnet gives bakers a tz5 manager key for transfers, delegation and staking, plus a separate tz6 consensus key that uses XMSS and a STARK-based aggregation step.
XMSS brings its own problem. Each key can produce only a limited number of signatures, so a baker has to track how many are left and rotate the key before it runs out. Nomadic Labs says the tutorial sizes the key for 131,071 signatures, roughly three days of baking. Rotation is manual, and the Octez software does not yet warn bakers as the budget runs down. Nomadic Labs calls this exactly the kind of constraint Quantumnet exists to expose.
What works on Quantumnet now
According to the Agora post, participants can run a node and bake, delegate and stake, send transactions with post-quantum keys, produce and attest DAL slots, and use a block explorer, a faucet, a public RPC endpoint and a metrics dashboard. Blocks come every 6 seconds and a cycle lasts 10 minutes. Nomadic Labs budgets about 45 minutes for the full baker tutorial.
Some things are deliberately missing:
- no rollup is running on the public network yet, although the plumbing for one exists;
- DAL bandwidth is kept modest while the team measures the system;
- the network topology is tightly controlled, unlike mainnet;
- hardware requirements are still being measured.
Later iterations may start from a fresh chain. Balances, delegate registrations and history are not expected to carry over, and Nomadic Labs tells testers to use throwaway keys.
How this could reach Tezos mainnet
No date exists. The Spotlight post says components can be “strengthened and eventually brought forward as protocol amendment proposals,” which would go through Tezos’ on-chain governance like the 21 upgrades before them. The Agora post lists what future iterations should explore: higher DAL bandwidth, a more realistic network, better XMSS key management, stateful addresses that would let an existing account add a post-quantum key without moving funds, and a rollup for post-quantum private transactions.
Tezos is not alone. Shielded Labs is working on post-quantum cryptography for Zcash with its Epoch project. For the standards behind ML-DSA, see what post-quantum cryptography is, and for why projects are starting years ahead of any capable quantum computer, see what Q-Day means.
