The short answer
The explorer software is the easy part and takes an evening. The node underneath is the real cost: over 700 GB and several days for Bitcoin with a transaction index, multiple terabytes for an Ethereum archive node. Self-hosting makes sense if you already run a node, if your lookups are sensitive, or if you need unlimited queries. For occasional checking it is the wrong trade.
Why anyone does this
Three reasons, and they are worth separating because they justify different amounts of effort.
Privacy. Every hosted explorer learns that someone at your IP address is interested in the addresses you looked up. Run your own and nobody learns anything, because there is no operator. This is the reason that cannot be achieved any other way — Tor helps considerably, and it does not eliminate the disclosure, it relocates it.
Rate limits. Free APIs throttle, and they are right to. Your own instance has no limit beyond your hardware, which for analysis work changes how you operate — you stop batching queries to be polite and simply ask.
Independence. No outage that is not yours, no key to expire, no terms of service to change, no company to be acquired and shut down. If something you depend on runs on someone else's goodwill, self-hosting converts that into infrastructure you control.
There is also a fourth reason people give that we would treat carefully: verification. Running the explorer yourself does mean the index is yours, but if you are trusting a third-party RPC rather than your own node, you have moved the trust rather than removed it.
Esplora, for Bitcoin
Esplora is Blockstream's open-source explorer and the most practical route to a self-hosted Bitcoin explorer. It is what runs blockstream.info, and the API shape has become a de facto standard — mempool.space implements a compatible surface, so code written against one works against the other.
What you need is a Bitcoin Core node with txindex enabled. That is the cost. A default node prunes data it no longer needs; an explorer node cannot, because someone will ask about a transaction from 2013. With the index, budget comfortably over 700 GB, and use an SSD — spinning disk will work and the initial sync becomes a multi-week exercise rather than a multi-day one.
The Esplora indexer then builds its own database on top, adding meaningfully to that figure. Initial sync and index construction takes days, not hours, and is largely unattended once started.
Once running, resource use is modest. A machine that can hold the data comfortably serves a household without difficulty, and the API you get is the same one this site queries.
The alternative is mempool.space's full stack, which adds the fee visualisation and the mempool views. It needs the same node plus an Electrum-protocol server, and the project publishes Docker images. If you want the interface rather than only the API, this is the better choice.
Blockscout, for EVM chains
Blockscout is the open-source EVM explorer that most rollups deploy on day one, and self-hosting is the use case it is actually built for. The stack is an indexer, a PostgreSQL database and a Phoenix web application, with Docker Compose configurations published for common setups.
You supply an archive node for the chain, and that is where the cost lives. An Ethereum mainnet archive node is a multi-terabyte commitment on fast storage — this is a server rather than a spare laptop, and the sync is measured in weeks rather than days if you are building it yourself rather than restoring a snapshot.
For a young rollup the picture is completely different, and this is the important nuance. Base launched in 2023. Its history is short, its state is small, and an archive node for it is a fraction of the Ethereum commitment. The same applies to Linea, Scroll and most L2s.
So the honest advice splits. Self-hosting Blockscout for Ethereum mainnet is a serious infrastructure project. Self-hosting it for a rollup you care about is a weekend, and it is why rollups adopt this pattern themselves — the chain is small, the software is free, and the explorer is running before the first user arrives.
Otterscan, the light option for Ethereum
If your goal is private Ethereum lookups for yourself rather than running a public service, Otterscan is a far better fit than Blockscout and almost nobody knows it exists.
It is deliberately thin. Rather than building its own indexer, it uses Erigon's — Erigon is an Ethereum client designed around efficient storage and rich tracing, and it already exposes most of what an explorer needs through custom RPC methods. There is no separate index to build and keep in sync; the node is the index.
That makes Otterscan small enough to be auditable and fast enough to be genuinely pleasant. Queries hit local disk rather than a remote API, and the difference changes how you work.
The requirement is firm: Erigon, in archive mode for full history. That is still a multi-terabyte commitment and there is no way around it. But it is one commitment rather than a node plus a separate indexing database, which is a meaningful saving.
There is also no hosted version, deliberately — a hosted instance would reintroduce exactly the operator Otterscan exists to remove. You run it or you do not use it.
The chains where this is genuinely easy
Most self-hosting guides focus on Bitcoin and Ethereum, which are the two hardest cases. Several chains are far more approachable and it is worth knowing which.
Monero is the strongest case in the whole category. The chain is considerably smaller than Bitcoin's, a monerod node is realistic for an individual, and Onion Explorer runs against it. The privacy argument is also strongest here, since querying a public server about your own Monero transactions partially undermines the reason you are using Monero.
Litecoin is a lighter version of the Bitcoin case, running the same mempool.space stack against a much smaller chain.
Stellar's Horizon is designed to be operated by anyone and the ledger is small. Hedera's mirror node is similarly a first-class part of the platform rather than something bolted on.
Young rollups, as above, are tractable because their history is short.
If you want the experience of running your own explorer without committing a server to it, one of these is a better starting point than Bitcoin.
Should you actually do this?
Being direct, because most guides are not.
Yes, if you already run a node. The marginal cost is close to zero and the benefit is immediate. Adding Esplora to an existing Bitcoin node, or Otterscan to an existing Erigon instance, is unambiguously worth doing.
Yes, if your lookups are sensitive. Researchers, journalists, anyone whose interest in an address would itself be revealing, anyone handling significant holdings. The node cost is real and so is the exposure it removes.
Yes, if you need unlimited queries. For analysis work, removing rate limits removes an entire class of friction.
No, if you check a transaction once a month. Use Blockstream.info or Blockchair over Tor and get most of the privacy benefit for none of the cost. Both publish onion services.
No, if the hardware would be a stretch. A half-synced node on inadequate storage is worse than no node — it gives you stale answers you believe. If you cannot commit the resources, do not start.
The useful framing is that self-hosting is a proportionate response to a specific need rather than a general virtue. Knowing which need you have is the whole decision.
Questions people ask
How much disk space does a self-hosted Bitcoin explorer need?
Comfortably over 700 GB for Bitcoin Core with txindex, plus the explorer’s own database on top. Use an SSD — spinning disk turns a multi-day sync into a multi-week one.
Can I run a block explorer on a Raspberry Pi?
For Bitcoin with external SSD storage, yes, and it is a common setup. For an Ethereum archive node, no — that is server territory.
Is Otterscan easier than Blockscout?
Much. Otterscan is a thin front end over Erigon’s own RPC, so there is no separate index to build. The node is still the cost, but it is one commitment rather than two.
Do I need an archive node?
For Ethereum, yes, if you want full history. For Bitcoin the equivalent is a node with txindex enabled, which is considerably lighter than an Ethereum archive node.
Which chain is easiest to self-host an explorer for?
Monero, where the chain is small and the privacy argument is strongest, or a young rollup whose history is short. Both are far more tractable than Bitcoin or Ethereum mainnet.