This is a joint post from Nomadic Labs, Trilitech, and Functori.
We are pleased to introduce our next Tezos protocol upgrade proposal. Destination: Tallinn!
As usual, the “true” name of the proposal is its hash: PtTALLiNtPec7mE7yY4m3k26J8Qukef3E3ehzhfXgFZKGtDdAXu
The command to inject or upvote the proposal via the Octez client is:
octez-client submit proposals for <baking_key> PtTALLiNtPec7mE7yY4m3k26J8Qukef3E3ehzhfXgFZKGtDdAXu
The Tallinn proposal builds on features introduced in recent upgrades and focuses on lower latency, faster finality, leaner consensus, stronger security, and improved on-chain efficiency.
The main changes are:
- 6-second block time: Lower latency and faster finality for a smoother overall experience.
- All bakers attest every block (once at least 50% use tz4): Stronger security, less load on nodes, and more predictable staking rewards.
- Address Indexing Registry: Improved efficiency and lower storage costs for Michelson apps.
Below, we expand on the proposed changes. For more technical details and additional minor changes, see the Tallinn proposal’s changelog. Note that the proposal also introduces breaking changes, listed here.
6-second block time
Continuing the trajectory of previous upgrades – and taking another step on the Tezos X roadmap – the Tallinn proposal reduces Layer 1 block time from 8 seconds to 6 seconds.
Shorter block times mean lower latency and faster finality. With Tallinn, Tezos’ two-block finality on Layer 1 would be reduced from 16 seconds to just 12 seconds.
Layer 2 solutions such as Etherlink also benefit, as data publication for Layer 2 is secured through inclusion in Layer 1 blocks. Interchain bridging benefits in a similar way.
The 6-second block time is achieved without increasing hardware requirements for bakers. Keeping validation accessible to everyone is key to maintaining a broad validator set for genuine decentralization and censorship resistance.
For more information about how the 6-second block time has been tested, see this post.
All bakers attest every block (once at least 50% use tz4)
The recent Seoul upgrade made it possible to aggregate hundreds of attestations into a single signature in each block, using the BLS signature scheme behind tz4 addresses.
Besides enabling drastic on-chain efficiency gains, aggregation also makes it possible to switch to a new approach to attestation: Having all bakers attest to every block instead of a limited subset of bakers.
With aggregation, more attestations will not consume more block space, and having all bakers attest brings several benefits:
- Stronger security: All bakers participating in consensus for each block improves robustness and fault tolerance.
- Predictable rewards: Attestation rewards become directly proportional to baking power, eliminating randomness from attestation committee selection.
- Leaner consensus: Removing selection logic lightens the load on nodes, enabling even shorter block times and more streamlined validation.
The Tallinn protocol proposal implements this change, but the feature will only activate once at least 50% of bakers use tz4 addresses. This is to ensure a smooth transition with bakers setting the pace.
Once active, the feature remains enabled, even if participation dips later. Baking from tz1/tz2/tz3 addresses will still be possible.
We remind bakers that current Ledger hardware signers are unable to process tz4/BLS signatures fast enough for Tezos consensus needs. However, tz4-compatible alternatives have been introduced, including Tezos RPi BLS Signer, TezSign, and Signatory.
For more details on the switch to all bakers attesting, see this post.
Address Indexing Registry
The Tallinn proposal also introduces the Address Indexing Registry for Michelson apps.
The registry can greatly reduce contract storage footprint, especially for NFT contracts and other large ledgers, by eliminating redundant address data. Instead of each token contract storing full addresses repeatedly, each address is stored once and assigned a compact numeric ID, cutting storage needs by up to 100x for some contracts.
Existing contracts will need to be updated to make use of this feature, but doing so benefits both the contract owners and the network as a whole.
- Cost-efficiency: Large NFT collections and token contracts become much cheaper to operate.
- Network-wide efficiency: The network-wide storage footprint grows more slowly.
- Higher throughput: Smaller operation size means more operations fit in each block.
While the Tallinn proposal introduces this registry on Layer 1, the mechanism will also be available on Tezlink, the upcoming Layer 2 for Michelson apps.
To learn more about the Address Indexing Registry, see this post.
Next steps
We encourage developers, bakers, and ecosystem teams to test their applications and tools on the dedicated test network, Tallinnnet, which will be live soon. We are happy to answer questions and receive feedback in the Tezos Discord.
A release candidate for Octez v24.0, which contains the Tallinn protocol, as well as general improvements, will be published shortly.
The proposal period ends on November 29. Don’t forget to vote!
If adopted, the Tallinn proposal would activate in early 2026, which is shaping up to be a pivotal year for Tezos. The changes in Tallinn streamline Layer 1 to ensure a fast, secure, and efficient consensus layer underpinning the next-level blockchain experience envisioned with Tezos X.
As always, with no compromise on decentralization.
Let’s go to Tallinn!