Market cap $2.20 T −2.36 % over 24 hours24h volume $73.18 BBTC dominance 58.2 %ETH 10.0 %Fear & Greed 10 Extreme fearCoins 21,463

State channels explained: from Spilman’s 2013 idea to Lightning and Hydra

State channels explained: how parties lock funds, swap signed updates off-chain and settle disputes, why rollups won on Ethereum, and what still runs.

Advanced 21 min readUpdated Read first: Sidechains vs rollups: Polygon PoS, Arbitrum and security compared
A signed contract and a handshake beside a padlock, chain links and network nodes
Table of contents
  1. How state channels work: the blockchain as a judge, not a ledger
    1. Open, update, close: the life of a channel
    2. Payment channels and state channels are not the same thing
  2. From Spilman’s one-way channel to two-way payment channels
  3. Generalised state channels on Ethereum, paper by paper
    1. Sprites (2017): shorter lock-ups on long routes
    2. Perun (2017): virtual channels through a hub
    3. Counterfactual (2018): contracts that never need to be deployed
    4. ForceMove and Nitro (2018-2019): channels opened without touching the chain
  4. How a dispute is settled: the latest signed state wins
    1. Replacement or penalty
    2. Watchtowers: paying someone to stay awake
  5. Virtual channels and hubs: a network without a channel to everyone
  6. Why channels lost Ethereum’s scaling race to Plasma, then rollups
  7. What is still running in November 2026
    1. Lightning: the largest channel network by far
    2. Cardano Hydra heads: isomorphic state channels
    3. Yellow Network and Nitrolite: channels for trading
    4. The quiet ones and the pivots
  8. State channels vs rollups vs sidechains
  9. What a state channel means for your money
    1. Taxes and rules in the US and UK

On 20 April 2013, a developer named Jeremy Spilman posted a short message to the bitcoin-development mailing list. He described how a user could pay a service in tiny increments without sending a transaction for each one. The user locks a deposit on the chain once. After that, every payment is a new transaction, signed but never broadcast, that hands the service a little more of the deposit. The service keeps only the latest one and publishes it when the session ends. “The current Main-Net behavior is actually very well suited for this application,” Spilman wrote.

That is the seed of state channels: a way for a small group of people to do many operations together while the blockchain only sees the beginning and the end. Thirteen years later the same trick, made two-way, carries the Lightning Network, which counted 26,741 public channels on 14 November 2026. On Ethereum, state channels were the first big scaling hope between 2015 and 2018, then lost the race to Plasma and to rollups. They never disappeared.

The previous lesson, sidechains vs rollups, showed two ways of moving trust: to a set of validators, or to Ethereum itself through proofs. A state channel does something else. It moves trust to the participants: each one holds signed messages that a contract on the main chain will honour if the others misbehave. Below: how the machine works, the papers that turned payment channels into general ones, why Ethereum bet on rollups instead, and which channel systems still run in November 2026.

How state channels work: the blockchain as a judge, not a ledger

A state channel is an agreement between a fixed set of parties, enforced by a smart contract or a script on a layer 1 chain. The parties lock money (or the starting state of an application) in that contract. Then they stop talking to the chain. They exchange messages directly, each one describing the new state and signed by everyone. When they are done, one of them submits the final signed state and the contract pays out accordingly.

ethereum.org sums it up in one line: “State channels allow participants to securely transact offchain while keeping interaction with Ethereum Mainnet at a minimum” (ethereum.org, state channels page). In the simplest case only two transactions ever touch the chain, one to open and one to close, however many updates happen in between. The deposit also works as a bond: a party that tries to cheat risks the money it has locked.

The chain does not record the updates. It sits in the background like a court that is only called when the parties disagree. That is why an update between parties is final the moment all of them have signed it. Nobody waits for a block, a sequencer or a proof. The cost is that everyone must keep their own copy of the latest signed state, because the chain does not.

Open, update, close: the life of a channel

  1. Open: the parties send their deposits to the channel contract and co-sign an initial state, for example “Alice 1 ETH, Bob 1 ETH, version 0”.
  2. Update: each change is a new state with a higher version number (a nonce), signed by every party. Old states are not deleted, they are simply outranked.
  3. Close together: if everyone agrees on the last state, they sign a closing message and the contract pays out at once.
  4. Close alone: if one party disappears or refuses, the other posts the latest state it holds. A challenge period opens during which anyone in the channel can show a state with a higher version. When the period ends, the highest version wins.

Take a chess game as an example, with invented figures. Alice and Bob each lock 1 ETH in a contract that knows the rules of chess. They play 40 moves off-chain: each move is a signed state numbered 1 to 40. Bob is checkmated and stops answering. Alice posts state 40 to the contract, the timer starts, Bob cannot produce a state 41 because none exists, and the contract sends 2 ETH to Alice. Forty moves, two on-chain transactions.

Two people beside a vault door, a signed document and isometric blocks
Two deposits, forty signed states, and only two transactions on the chain.

Payment channels and state channels are not the same thing

A payment channel tracks one thing: who owns how much of the deposit. Every update moves balance from one side to the other. A general state channel tracks anything a contract can store: the position of a chess board, the hands of a card game, the terms of a bet, an order between two traders. Payment channels are the special case, and historically they came first.

The difference matters for the contract. A payment channel contract only needs to check signatures and sums. A general state channel contract needs the rules of the application, so that it can tell whether a disputed state follows from the previous one. That is why general channels appeared on Ethereum, with its programmable contracts, and payment channels on Bitcoin, whose script language is deliberately limited.

From Spilman’s one-way channel to two-way payment channels

Spilman’s design, discussed in the thread “Anti DoS for tx replacement” opened by Mike Hearn on 16 April 2013, needed no change to Bitcoin. The payer funds an output that needs both signatures to spend. Before signing it, the payer gets the service to sign a refund transaction with a lock time in the future, so the deposit comes back if the service vanishes. Then each payment is a new version of the spending transaction, a little more generous to the service each time. The service broadcasts the most generous one before the refund becomes valid.

The design has two built-in limits. Money flows in one direction only, from payer to receiver. And the channel expires: the receiver must close before the refund’s lock time, or lose the payments. bitcoinj, a Java wallet library, implemented these channels in 2013. The CLTV opcode (BIP65), activated on Bitcoin in December 2015, made them simpler and removed a weakness linked to transaction malleability, but they stayed one-way and time-limited.

Two-way channels came next. In 2015 Christian Decker and Roger Wattenhofer of ETH Zurich proposed duplex micropayment channels, built from two one-way channels pointing in opposite directions and refreshed with relative lock times. In January 2016 Joseph Poon and Thaddeus Dryja published version 0.5.9.2 of the Lightning Network paper, with channels that can run indefinitely and in both directions.

Poon and Dryja’s channels are the ones Bitcoin uses today. Each side holds a commitment transaction spending a 2-of-2 funding output, old commitments are revoked by handing over a secret, a cheater who publishes a revoked state loses the whole channel, and hash time-locked contracts (HTLCs) let payments hop across several channels. Our lesson on the Lightning Network covers that machinery, routing and watchtowers in detail; this lesson keeps to what is general to all channels. Statechains, despite the name, are a different Bitcoin design, described in our map of Bitcoin layer 2s and sidechains.

Generalised state channels on Ethereum, paper by paper

Once Ethereum offered contracts that could hold arbitrary rules, researchers asked whether a channel could run any application beyond payments. Between 2017 and 2019 a series of papers answered yes, each removing one limit of the previous design. The dates below are those of the first public versions.

DateDesignAuthorsWhat it added
19-20 Apr 2013Spilman channelJeremy SpilmanOne-way payment channel: deposit, refund with lock time, receiver publishes the last payment
2015Duplex micropayment channelsChristian Decker, Roger WattenhoferTwo-way payments from two one-way channels
Jan 2016Lightning paper (v0.5.9.2)Joseph Poon, Thaddeus DryjaRevocable two-way channels, HTLC network
19 Feb 2017SpritesAndrew Miller, Iddo Bentov and othersShorter collateral lock on long routes, deposits and withdrawals without closing
28 Jun 2017PerunStefan Dziembowski, Lisa Eckey, Sebastian Faust, Daniel MalinowskiVirtual payment channels through a hub
8 Apr 2018General State Channel NetworksStefan Dziembowski, Sebastian Faust, Kristina HostakovaFormal specification of networks running any contract
6 Jun 2018PisaPatrick McCorry and othersPaid, accountable watchtowers
12 Jun 2018CounterfactualJeff Coleman, Liam Horne, Li XuanjiCounterfactual instantiation, a multisig as the only on-chain piece
27 Feb 2019Nitro protocolTom CloseChannels funded and closed without on-chain transactions, n parties, virtual channels
9 Mar 2020HydraManuel Chakravarty and others (IOHK)Isomorphic channels that reuse Cardano’s own transaction format

Sources: the papers on arXiv and the IACR ePrint archive, the bitcoin-development mailing list and the Lightning paper; dates are the first posted versions.

Sprites (2017): shorter lock-ups on long routes

The Sprites paper by Andrew Miller, Iddo Bentov, Ranjit Kumaresan, Christopher Cordi and Patrick McCorry looked at a payment crossing many channels. In a Lightning-style route each hop gets its own deadline, and the deadlines have to step down hop by hop, so the collateral of a long route can stay frozen for a time that grows with the length of the route multiplied by the dispute delay. Sprites uses one shared contract on the chain to record the payment’s secret, so all hops share a single deadline and the lock-up grows with the route length plus the delay instead. It also let parties add or withdraw funds without stopping the channel.

Perun (2017): virtual channels through a hub

Perun, by Stefan Dziembowski, Lisa Eckey, Sebastian Faust and Daniel Malinowski, introduced virtual payment channels. If Alice and Bob both have a channel with Ingrid, they can build a channel between themselves on top of those two, and then pay each other without Ingrid taking part in each payment. Ingrid only acts when the virtual channel opens and closes. The authors built a proof of concept on Ethereum. A follow-up, General State Channel Networks (Dziembowski, Faust and Hostakova, April 2018), gave the first full specification, with security proofs, of channel networks that run arbitrary contracts.

Counterfactual (2018): contracts that never need to be deployed

Jeff Coleman, Liam Horne and Li Xuanji of L4 presented Counterfactual in June 2018. Their idea, counterfactual instantiation, binds the parties to a contract just by signing messages about it: the contract would be deployed on Ethereum only in a dispute. The only on-chain object per channel is a multisig wallet, which helps privacy and lets the applications inside the channel be upgraded. Their “metachannels” play the same role as Perun’s virtual channels.

ForceMove and Nitro (2018-2019): channels opened without touching the chain

Magmo’s ForceMove protocol (2018) handled turn-based games: if a player stops responding, the other can force a move on the chain, and the silent player must answer within a time limit or lose. Tom Close’s Nitro protocol paper of February 2019 built on it. In Nitro, a channel can be funded from another channel, so new channels can be opened and closed without any on-chain transaction, fewer deposits sit idle, and the design extends to more than two parties and to virtual channels between three. Nitro became the basis of the statechannels.org project.

How a dispute is settled: the latest signed state wins

Every channel design answers the same question: what stops Bob from closing with an old state that suited him better? The answer is the version number. Each state carries a counter that only goes up, and the contract accepts a newer state over an older one. A unilateral close is therefore never instant. It opens a challenge window during which the other parties can present a higher version. ethereum.org describes it this way: newer states override older ones, joint closes settle at once, and exits by a single party are delayed.

The logic is close to the double-spend problem that blockchains solve with ordering: the chain cannot know which of two signed states is real, so the protocol gives it a rule it can check mechanically. The “proof” in a channel dispute is nothing more than a newer state carrying everyone’s signature.

Replacement or penalty

Generalised Ethereum channels mostly use replacement: if Bob posts state 30 and Alice shows state 40, state 40 simply takes its place and Bob loses nothing but the gas he spent. Lightning uses a penalty: a party that publishes a revoked commitment can lose all the funds in the channel. Penalties deter cheating more strongly but make a mistake, such as restoring an old backup, very costly. The eltoo design proposed by Christian Decker, Rusty Russell and Olaoluwa Osuntokun in 2018 would bring replacement to Bitcoin, but needs a new signature flag in Bitcoin’s rules.

Cardano’s Hydra uses the same idea under another name. When a head closes, a contestation period starts, and an honest party “can use signed snapshots to contest an outdated close, provided it monitors Cardano within the contestation period”, according to the Hydra documentation.

Watchtowers: paying someone to stay awake

A challenge window only protects you if you are watching. A phone that stays off for a week could miss it. Watchtowers are services that watch the chain on a user’s behalf and react to a stale close. The Pisa paper of June 2018, by Patrick McCorry, Surya Bakshi, Iddo Bentov, Andrew Miller and Sarah Meiklejohn, treated the requirement to stay online as the weak assumption of state channels and proposed a paid watchtower that can be held accountable if it fails. ethereum.org notes the trade-off: watchtowers help with liveness but add a cost, and a new party to trust.

Virtual channels and hubs: a network without a channel to everyone

Opening a channel with every person you deal with would cost a deposit and two transactions each time, which defeats the purpose. Two answers emerged. Lightning routes payments across existing channels with HTLCs: each intermediary forwards the payment in exchange for a fee and is involved in every single payment. Virtual channels, as in Perun, Counterfactual’s metachannels and Nitro, take a different path: the intermediary helps set up a new channel on top of existing ones, then steps aside until it closes.

Both lead to the same economic pull. The parties best placed to connect everyone are those with many channels and a lot of capital, so networks tend to form around hubs. A hub cannot steal funds in a well-built design, but it can refuse to route, it sees a lot of traffic, and its capital is tied up in proportion to the volume it carries. And none of this changes the basic rule: every channel, real or virtual, has a fixed list of participants decided when it opens.

Why channels lost Ethereum’s scaling race to Plasma, then rollups

In 2017 and 2018 Ethereum had several funded channel projects: Raiden for payments, Counterfactual, Magmo, Celer, Connext and others. By 2020 the ecosystem had changed course. The reasons come out of the design itself, and ethereum.org lists most of them in its limitations section.

  • Fixed participants: a channel only works between the parties who opened it. You cannot pay or trade with someone who is not in your channel or reachable through the network.
  • Locked capital: deposits sit in the contract for the whole life of the channel, and a hub has to lock enough to cover every user it serves.
  • Liveness: each participant, or a watchtower paid by them, must be online to answer a dispute within the challenge window.
  • Griefing: a counterparty can post an old state or go silent, forcing the honest side to spend gas and wait.
  • Data: users must store the latest signed state themselves. Losing it can mean losing money.
  • No open, shared state: a smart contract holding reserves of two or more tokens, supplied by users, that a decentralized exchange uses to execute trades automatically at a price set by a formula.">pool on an exchange or a lending market is a set of balances that any stranger can use at any time, with no fixed owner. A channel among a fixed group cannot host that.

For DeFi the last point mattered most. Decentralised finance grew in 2020 around shared contracts like trading pools and lending markets, used by anyone at any moment. Channels suit repeated interactions between known parties. They do not suit an open market.

Plasma came first as the alternative. Joseph Poon and Vitalik Buterin’s Plasma paper, a working draft dated 11 August 2017, proposed child chains that post only commitments to Ethereum, with an exit game for users to leave with their funds. It removed the fixed-participants problem but kept the data problem. In a 2023 post Buterin wrote that “Plasma has since largely been superseded by rollups”, mainly because of client-side data storage costs and limits that made it hard to go beyond payments (vitalik.eth.limo, 14 November 2023).

Rollups solved the data problem by posting transaction data to Ethereum itself, so anyone can rebuild the state and nobody has to keep private copies. On 2 October 2020 Buterin concluded that “the Ethereum ecosystem is likely to be all-in on rollups (plus some plasma and channels) as a scaling strategy” (Ethereum Magicians). Optimistic rollups defend that data with fraud proofs, ZK rollups with validity proofs, and our guide to what a layer 2 is shows how the stages of decentralisation are rated. ETHBEAT, formerly L2BEAT, tracks rollups and validiums; no channel system appears in its tables as of 14 November 2026.

An hourglass and a watching eye beside a warning sign and a server rack
Challenge windows and watchtowers: the liveness cost that pushed Ethereum towards rollups.

What is still running in November 2026

Channels did not vanish when Ethereum moved to rollups. Several projects went quiet or turned into bridges; a handful are live. The status below was checked on public repositories, documentation and statistics on 14 November 2026.

SystemChainStatus as of 14 November 2026
Lightning NetworkBitcoin26,741 public channels, 13,847 nodes, about 3,244 BTC of public capacity (mempool.emzy.de)
Hydra headsCardanohydra-node 2.4.2 released 22 October 2026; first mainnet-usable release 0.10.0 on 11 May 2023
Yellow Network (Nitrolite)Ethereum and othersMainnet launch announced in a sponsored press release in March 2026; ERC-7824 still an open pull request
Perun (go-perun)SeveralOpen-source framework, repository updated 27 October 2026
RaidenEthereumLast release v3.0.1 on 13 January 2022; development quiet since 2022
Nitro (statechannels.org)EthereumMain monorepo archived (last update October 2022); go-nitro described as pre-production
Connext, now EverclearEthereum and L2sState channel code archived; rebranded as a clearing layer for cross-chain settlement in June 2024
CelerSeveralMoved from a state channel network to cross-chain messaging and the cBridge bridge

Lightning: the largest channel network by far

As of 14 November 2026, mempool.emzy.de, a community instance of the mempool.space software, counted 26,741 public Lightning channels between 13,847 nodes, with a combined public capacity of about 3,244 BTC. mempool.space’s own Lightning feed stopped updating on 30 August 2026, when it showed 32,674 channels and about 3,794 BTC. These figures only count announced channels: private channels between wallets and their providers are not included. The Lightning lesson analyses the trend.

Cardano Hydra heads: isomorphic state channels

Hydra was described in the 2020 paper “Hydra: Fast Isomorphic State Channels” by Manuel Chakravarty, Sandro Coretti, Matthias Fitzi, Peter Gazi, Philipp Kant, Aggelos Kiayias and Alexander Russell. “Isomorphic” means a head reuses Cardano’s own transaction format and validation rules off-chain, so code written for Cardano can run inside a head. In the words of the documentation, “A fixed group of parties opens an empty Hydra head on Cardano”. Release 1.0.0 arrived on 8 October 2025, and 2.0.0 on 2 April 2026 removed the old commit phase: heads now open empty and funds are added by deposits.

Hydra’s best-known public test was a Doom tournament with a $100,000 prize pool, run by IOG, Cardano’s main developer, and the Hydra team in December 2024. The organisers said the game reached about one million transactions per second at its peak across many heads running in parallel. That was a demonstration, not the throughput of Cardano: CryptoSlate reported critics’ point that those transactions were not settled on Cardano’s mainnet. No public figure shows how much value Hydra heads hold in production.

Yellow Network and Nitrolite: channels for trading

Yellow Network builds a trading and clearing network on state channels using Nitrolite, which its repository describes as a simplified version of Nitro. According to a sponsored press release carried by Decrypt in March 2026, the protocol went live on Ethereum mainnet, with node operators locking YELLOW tokens to run “Clearnode” servers. Those are the company’s own statements; no independent usage figure exists. Its proposed standard, ERC-7824, is not an adopted ERC: the pull request opened on 22 November 2024 was still open on 14 November 2026.

The quiet ones and the pivots

Raiden, an Ethereum payment channel network whose code repository dates from January 2016, published its last release, v3.0.1 “Coruscant”, on 13 January 2022. Its repository has seen little since, and no shutdown notice exists. The statechannels.org monorepo was archived in 2022, and its Go implementation, go-nitro, still warns that it is “pre-production software”. Perun remains active as a chain-agnostic framework in Go.

Connext started as a state channel network (its Indra and Vector code, from 2018 and 2020, is now archived), built a bridge, then rebranded as Everclear in June 2024 to run a clearing layer for intent-based transfers between chains, as The Block reported at the time. Our guide to cross-chain liquidity, bridges and intents explains that model. Celer published its whitepaper on a generalised state channel network in September 2018, then turned its State Guardian Network into the backbone of cross-chain messaging and the cBridge bridge.

What remains are the niches where channels fit naturally: frequent payments between the same parties (Lightning, streaming or pay-per-use services), turn-based games, and trading between a client and a known counterparty.

State channels vs rollups vs sidechains

The three families answer different questions. This table compares them on what a user feels: who must be online, who holds the money and how to leave.

State channelRollupSidechain
Who must be online for safetyEach participant, or a watchtower, during any challenge windowNobody: data on Ethereum lets users exit; the sequencer only matters for livenessThe sidechain’s validators
Who holds the fundsA contract or 2-of-2 output controlled by the participants’ signaturesThe rollup’s bridge contract on EthereumThe bridge, run by validators or a multisig
How to exit alonePost the latest signed state and wait out the challenge periodWithdraw through the bridge: about 6.4 days on Arbitrum One, shorter with ZK proofsThrough the bridge operators
Where the data livesWith each participantOn Ethereum (blobs)On the sidechain
Who can joinOnly the parties of the channelAnyoneAnyone
Finality between usersInstant once all have signedFast confirmation, final on Ethereum laterSet by the sidechain’s consensus
On-chain costOne transaction to open and one to close; updates are freeA small fee per transactionA small fee per transaction
Best fitRepeated payments or games between known partiesGeneral apps, DeFi, open marketsApps that accept the validators’ trust

A channel is the only option where nobody outside the group can stop or reorder your updates, and where an update is final with no block at all. It is also the only one where your own laziness can cost you money. Rollups trade a little cost and some waiting for open participation and data that nobody has to keep privately.

What a state channel means for your money

Most people meet channels through a Lightning wallet or an app built on a channel network. Before depositing, a few checks matter more than the fees.

  • Lost state: if your device loses the latest signed state, you may be unable to prove what you are owed. Penalty-based channels add a second risk, since publishing an old backup can cost the whole balance.
  • Being offline: a counterparty can try to close with an old state while you are away. Know how long the challenge window is and whether a watchtower covers you.
  • Griefing and timeouts: a counterparty or a hub can go silent, leaving your funds locked until the timeout and forcing you to pay fees to close.
  • Contract bugs: the adjudicator contract holds everything. Check whether its code is open and audited.
  • Custody: some apps show channels on screen but hold the keys themselves. That is a custodial service, whatever the technology underneath.

Before depositing, ask whether you can leave alone, with only what is on your device, if every other party disappears tomorrow? If the answer is no, you are trusting someone, and you should know whom.

Taxes and rules in the US and UK

In the US and the UK, rules on crypto services attach to whoever holds customers’ money, not to the protocol underneath, so a custodial channel app is treated as a crypto service like any other. For taxes, the IRS treats paying for goods or services with digital assets as a disposal, so a payment through a channel can create a gain or loss in dollars just like an on-chain one; HMRC applies the same logic to Capital Gains Tax. Our guides to crypto taxes in the United States and crypto taxes in the United Kingdom cover record keeping.

Channels will not carry Ethereum’s open markets, but they remain the fastest way for a known group to settle many small operations. The next lesson of this path, on modular blockchains and appchains, looks at the opposite approach: splitting a blockchain into specialised layers that anyone can use.

Sources

  1. ethereum.org: State channels ethereum.org
  2. Jeremy Spilman, bitcoin-development mailing list, 20 April 2013 mirror.b10c.me
  3. Miller et al., Sprites and State Channels (arXiv 1702.05812) arxiv.org
  4. Dziembowski et al., Perun: Virtual Payment Hubs over Cryptocurrencies (ePrint 2017/635) eprint.iacr.org
  5. Coleman, Horne, Xuanji: Counterfactual, generalized state channels (ethresear.ch) ethresear.ch
  6. Tom Close, Nitro protocol (ePrint 2019/219) eprint.iacr.org
  7. Hydra Head protocol documentation hydra.family
  8. Vitalik Buterin, A rollup-centric Ethereum roadmap (2 October 2020) ethereum-magicians.org
  9. mempool.emzy.de Lightning statistics mempool.emzy.de
  10. ethereum/ERCs pull request 728: Nitrolite State Channels github.com
  11. Poon and Buterin, Plasma: Scalable Autonomous Smart Contracts (working draft, 11 August 2017) plasma.io
  12. Vitalik Buterin, Exit games for EVM validiums: the return to Plasma (14 November 2023) vitalik.eth.limo
  13. McCorry et al., Pisa: Arbitration Outsourcing for State Channels (ePrint 2018/582) eprint.iacr.org
  14. Dziembowski, Faust, Hostakova, General State Channel Networks (ePrint 2018/320) eprint.iacr.org
  15. Chakravarty et al., Hydra: Fast Isomorphic State Channels (ePrint 2020/299) eprint.iacr.org
  16. Decrypt press release (sponsored): Yellow Network launches on Ethereum mainnet, 16 March 2026 decrypt.co
  17. Bitcoin Wiki: Payment channels en.bitcoin.it
Coindaro Editorial Team

Who wrote this article

Editorial team

Guides, reviews and glossary checked against primary sources

  • Crypto education
  • Exchange reviews
  • Wallets
  • Glossary

The Coindaro editorial team researches and writes every article, review and guide published on Coindaro. Each fact is checked against primary sources such as official documents, regulators and project documentation, and every page lists its sources. We cover crypto markets, technology and regulation for readers in the United States and Europe, independently and without paid placements.

Keep learning