What is archive node?

A node that keeps every historical state of the chain, not just the current one, so it can answer what a balance or contract variable was at any past block.

Ordinary full nodes discard intermediate states because they can be recomputed, which makes archives far larger and rarer. They exist to serve explorers, indexers, accounting tools, and analytics that need history on demand. When an API tells you a historical balance query costs more or is unavailable on a free plan, an archive node is why.

Cost is why so few exist, and it is why historical analysis of a chain concentrates around a small number of operators who can afford to keep the data.

How it works

A blockchain stores transactions permanently, but state, meaning every account balance and every contract variable, is a separate structure that changes with each block. A full node verifies every block and keeps the current state plus a short recent window, pruning older snapshots because they can always be rebuilt by replaying from genesis.

An archive node skips the pruning. It retains the state at every historical block, so a query like "what was this address holding at block 12,000,000" is a lookup rather than a multi-day replay.

Storage requirements are the practical constraint. Archive data grows monotonically with chain history and, on busy chains, runs to many terabytes, so operators need fast solid-state storage rather than cheap bulk disks.

Archive nodes add nothing to consensus. They validate exactly what a full node validates and have no extra authority. The difference is purely what they keep.

Example

Illustrative: your tax software needs your token balance at the end of each of 36 months across three chains. That is 108 historical state queries. An ordinary full node cannot answer them without replaying the chain; an archive node answers each in milliseconds. This is why tax and portfolio tools sell subscriptions and why free tiers of hosted node providers usually block historical queries.

Why it matters when you buy

You never run one, but you depend on them constantly. Every explorer page showing a past balance, every tax importer reconstructing your cost basis, and every analytics dashboard sits on archive data. When a tool reports gaps in your history or cannot price an old transaction, an archive limitation is often the reason. See the crypto tax basics guide for why that history matters at filing time.

Related terms

• full node — validates everything but prunes old state
• node — the general term for a network participant
• rpc provider — the companies that rent access to nodes
• light client — the minimal opposite of an archive
• indexer — turns chain data into queryable records
• block explorer — the front end most people use

Questions

Do I need an archive node to check my own history?

No. Explorers and portfolio tools query one on your behalf. You would only run one to avoid depending on a third party or to do bulk analysis.

Why is historical data a paid feature?

Because storing every past state costs far more than storing the current one, and the queries are heavier. Providers price that separately from ordinary current-state calls.

Is archive data ever lost?

The transactions are not, since any node can replay them. What can disappear is the convenience of instant historical queries if the operators serving them stop, which is a real dependency for tooling rather than for the chain.