Every protocol upgrade is a chance to refine the foundations of Tezos. In our upcoming protocol upgrade proposal, one of the improvements is the Address Indexing Registry — a mechanism that greatly improves how token and NFT ownership is stored on-chain.
Up until now, every contract had to store full addresses repeatedly in its ledger, even when the same account appeared thousands or millions of times in the whole ledger. With the new registry, an address is registered once and assigned a compact numeric ID, taking up just 3-5 bytes instead of 22 bytes. Contracts then store (token_id → address_id) instead of (token_id → address).
Stored addresses, such as internal ledgers in FA2 contracts, are a major factor for contract storage consumption. The Address Indexing Registry and related optimizations make it possible to reduce contract storage by as much as 50–100x for contracts that are designed to make full use of this functionality.
Why it matters
-
More cost-efficient applications: Large NFT collections, token contracts, and bulk transfers become much more accessible on Layer 1.
-
Scalable by design: The heavy lifting is shared at protocol level: addresses are stored once in a global registry rather than repeated across every contract.
-
Network-wide efficiency: Adopting this mechanism for smart contract design will slow the growth in storage footprint, benefitting both the contract storage cost and the Tezos blockchain as a whole.
-
Higher throughput: Cutting the per-operation storage footprint means more operations can be included in each block.
This reduces the per-contract storage load, distributing it across the network in a consistent and optimized way.
Technical details
For builders who want the specifics:
-
Global registry: Any address can be registered once, receiving a unique numeric ID (nat).
-
Contract integration: Instead of
big_map token_id → address, new contracts can now usebig_map token_id → nat. The mapping fromnat → addressis maintained by the protocol. -
Michelson semantics: New instructions provide registration (
INDEX_ADDRESS) and lookup (GET_ADDRESS_INDEX). Gas and storage costs are deterministic. -
Efficient storage: The registry reuses Tezos’ context storage system (already Patricia trie–based) to ensure efficient insertions, lookups, and low overhead.
-
Compatibility: Existing contracts remain unaffected. Adoption is optional and can be incremental.
In practice, this means significantly lighter internal ledgers for FA2-style contracts that make use of the global registry, lower storage costs, and the ability to sustain much higher operation counts per block.
Unlocking further optimisations
The Address Indexing Registry is also a stepping stone. By avoiding repeated address data, it enables in-contract storage optimisations — as demonstrated in the FA2 demo project.
Combined with existing Tezos features like off-chain metadata, views, and other composable contract patterns, developers can achieve dramatically lower storage costs for tokens and NFTs than previously possible on Layer 1.
Path to rollup & TezLink
If the proposal is adopted, this mechanism arrives first on Layer 1, but it will also be part of TezLink and Tezos’ canonical rollup, which will further amplify cost savings and performance gains for Michelson smart contracts.
By distributing storage more intelligently across the network, and pairing lighter ledgers with faster 6-second blocks, Tezos is being prepared for a new wave of scalable, cost-efficient dApps.
