Blockchair: the explorer that lets you ask questions

The query layer is the whole point

Type a transaction hash into any explorer and you get that transaction. Useful, and entirely unremarkable. Blockchair is the only free explorer where you can instead say: show me every Bitcoin transaction confirmed yesterday with more than fifty inputs, sorted by fee rate, and count them.

It does this by exposing its index as a queryable surface rather than a set of fixed pages. Every field it has stored — block height, time, input count, output count, fee, fee per byte, size, weight, whether the transaction signalled replace-by-fee — can be filtered on, sorted by, and in many cases aggregated.

The same structure appears in the URL and in the API, which is the design decision that makes the whole thing work. A query you build by clicking around the web interface can be lifted straight into a script:

GET https://api.blockchair.com/bitcoin/transactions
  ?q=time(2026-09-16..2026-09-17),input_count(50..)
  &s=fee_per_kb(desc)
  &limit=100

There is no meaningful equivalent on Etherscan or mempool.space. To answer that question with either of them you would pull raw blocks and filter locally, which for a week of Bitcoin history is a real piece of work rather than a two-minute task.

Twenty chains, one schema

Blockchair indexes Bitcoin, Bitcoin Cash, Ethereum, Litecoin, Dogecoin, Dash, XRP, Stellar, Monero, Cardano, Zcash, eCash, Groestlcoin and several more. The coverage is not the widest available — Tokenview claims more, and OKLink covers more EVM networks — but the consistency is the thing nobody else matches.

Because the schema is shared, a query written against Bitcoin generally works against Litecoin or Dogecoin by changing one path segment. For anyone comparing behaviour across UTXO chains, that is the difference between an afternoon and a fortnight.

It handles the awkward chains honestly too. On Monero it indexes what is publicly visible — block structure, transaction existence, ring sizes — and does not pretend to amounts or participants, because those are cryptographically hidden. Explorers that imply otherwise on Monero are showing you a guess.

A keyless API that still works

We tested this with live requests rather than reading the pricing page, because the gap between the two is where most API comparisons go wrong. A plain request to the stats endpoint returned a full result with no key, no registration and no email address:

curl https://api.blockchair.com/bitcoin/stats

It also returns Access-Control-Allow-Origin: *, which means it can be called directly from a browser. That is a bigger deal than it sounds. Etherscan and Blockscout both require a key, and a key cannot be safely put in front-end code, so both force you to run a server. Blockchair does not. The lookup tool on this site used exactly this property until the rate limits below made it unusable, which is the whole story of this API in one sentence.

The dashboards endpoints are the ones worth learning. A single call returns an address with its balance, transaction count, first and last activity, and a page of transactions — where most APIs would need three or four round trips.

Privacy: better than most, still a third party

Blockchair scores well here by the standards of hosted explorers, for three concrete reasons rather than any promise.

It runs an onion service, so a lookup can be made over Tor without exiting to the clear web. It renders usable content without JavaScript, which removes an entire class of browser fingerprinting. And it publishes bulk database dumps, which means the heaviest users — researchers, analysts, anyone whose query pattern would itself be revealing — can take the data and work locally.

That last point is the one people overlook. If your queries are sensitive, the answer is not a better privacy policy, it is not sending the queries at all. Blockchair is one of very few explorers that makes that practical without you running a node.

What it cannot offer is verifiability. It is a hosted service with closed infrastructure, and you are trusting an operator. If that is unacceptable, the honest answers are Otterscan against your own node, or a self-hosted Esplora instance. Our privacy guide sets out where each option actually lands.

Who it is actually for

If you have been given a transaction ID and told to check it, Blockchair is the wrong tool. It will answer you, but the page is dense, the terminology is unforgiving, and mempool.space will tell you what you need in a quarter of the time with a clearer picture of where you sit in the queue.

If you are trying to establish a pattern — how fee rates moved during a congestion event, how many addresses of a particular script type appeared in a period, whether two chains behaved similarly during the same window — there is nothing else free that comes close. Dune is more powerful for aggregate analysis but requires you to write SQL against its schema and meters your query credits; Blockchair answers a large class of questions from a URL.

And if you are building something on a static site or in a browser extension where you cannot hold a secret, a CORS-enabled keyless API is one of a handful of options that exist at all. That constraint is invisible until you hit it, and then it is the only thing that matters. Blockchair still technically qualifies — but on the allowance we measured, plan for it to stop answering, and decide now what your code does when it does.