When Sara’s 0.5 ETH transfer seemed to vanish, the answer did not come from a company helpdesk. It came from thousands of ordinary computers around the world, each holding an identical copy of Ethereum’s records, each able to confirm that her transaction was real. Those computers are called nodes, and they are the reason a blockchain works without a boss. No node, no blockchain. No blockchain, no crypto.

Yet most people who own crypto have never thought about what a node is, what it does all day, or why anyone would run one for free. This guide explains what is a blockchain node, the different types, how they keep a chain honest, and what it actually takes to run one yourself — including the costs, the hardware, and the rewards. If you need the basics first, our plain-English guide to blockchain technology explains blocks and chains from zero, and our guide to how blockchain works walks through the process step by step.

Executive Summary

• A blockchain node is a computer that runs the blockchain’s software. It stores a copy of the chain, checks new transactions, and talks to other nodes.
• Nodes are the reason blockchains are decentralised. No single node controls the chain, so no single failure or attack can stop it.
• There are several types. Full nodes, light nodes, archive nodes, validator nodes, and mining nodes each do a different job.
• Running a node is how you verify, not trust. Your own node checks every transaction itself instead of believing someone else’s server.
• Some nodes earn rewards, most do not. Validators and miners earn fees and block rewards; ordinary full nodes earn nothing but independence.
• You can run one at home. A Bitcoin or Ethereum full node needs a modest computer, a large hard drive, and a stable internet connection — not a data centre.

What Is a Blockchain Node, in One Sentence?

A blockchain node is any computer that runs a blockchain’s software, keeps a copy of its records, and follows the same rules as every other node on the network. Think back to the shared notebook from our beginner’s guide: ten friends each hold an identical copy, and a new page is added only when the group agrees. Each friend is a node. The notebook is the blockchain. The agreement is consensus.

That simple picture explains the whole system. A blockchain is not a place or a company. It is a network of nodes, each doing the same job independently. When you send a transaction, your wallet broadcasts it to a few nodes. Those nodes check it — is the signature valid, does the sender have the funds, is the format correct? — and pass it on to other nodes. Eventually a validator or miner includes it in a block, and every node adds that block to its copy. If one node lies about a transaction, the others ignore it, because their copies do not match. Our detailed explanation of blockchain technology covers how consensus and cryptography make that agreement work.

Node vs Miner vs Validator: Clearing Up the Confusion

Every miner and every validator is a node, but not every node is a miner or a validator. Most nodes simply store and check data. That is the job that keeps the network honest, and it is the job anyone can do from home.

How Nodes Talk to Each Other

Nodes connect to each other in a peer-to-peer network — there is no central server telling them what to do. When a node comes online, it connects to a handful of other nodes it already knows about, asks them for the latest blocks, and starts sharing what it has. New transactions spread across the network the way gossip spreads through a town: one node tells three, those three tell nine, and within seconds the whole network has heard.

This design is why blockchains are hard to shut down. There is no office to raid and no server to unplug. To stop Bitcoin, you would have to switch off tens of thousands of nodes in dozens of countries at the same time. To censor a transaction, you would have to control enough of the network to keep it out of every block — which is why the cost of attacking a large chain is measured in billions of dollars. Our guide to blockchain mining explains how miners and validators are chosen to write the next block, and our complete guide to crypto exchanges explains why serious platforms run their own nodes rather than trusting someone else’s.

What Happens When a Node Receives a Transaction

The checks are fast, automatic, and strict. First, the node checks the digital signature: did the sender really authorise this? Second, it checks the balance: does the sender actually have the funds? Third, it checks the format: is the transaction well-formed and does it follow the chain’s rules? Fourth, it checks for double-spending: has this money already been spent in another transaction? If any check fails, the node drops the transaction and does not pass it on. If all checks pass, the node stores it in its mempool — a waiting room for unconfirmed transactions — and forwards it to its neighbours. A transaction that reaches most nodes within seconds is a transaction the network has accepted as valid.

The Types of Blockchain Nodes

Not all nodes do the same work. The type depends on how much data the node stores and what role it plays in creating blocks. Choosing the right type matters: a light node fits in a phone, while an archive node needs terabytes of storage and a serious machine.

Full Nodes

A full node stores the entire blockchain and checks every transaction and every block against the rules, from the very first block to the latest one. This is the gold standard of independence: a full node trusts nobody. It verifies everything itself. Running a full node is also the best way to support the network, because each additional full node makes the chain harder to attack and easier for others to sync from. Bitcoin Core is the best-known full node software. A Bitcoin full node in late 2026 needs roughly 700 GB of disk space and a stable broadband connection, and the number grows every month as new blocks are added.

Light Nodes

A light node, also called an SPV node (Simplified Payment Verification), stores only block headers — small summaries of each block — instead of the full history. It asks full nodes for the details it needs, and it checks those details against the headers it holds. Light nodes are what most wallet apps use: they are fast, they fit on a phone, and they are good enough for everyday payments. The trade-off is trust. A light node depends on the full nodes it talks to, so it cannot verify everything on its own. For most users, that trade-off is worth it.

Archive Nodes

An archive node stores everything a full node stores, plus every historical state — the balance of every address at every block in history. Developers, block explorers, and analytics companies run archive nodes because they need to answer questions like “what was this wallet’s balance on 14 March 2021?” An Ethereum archive node can need several terabytes of fast SSD storage, which is why almost nobody runs one at home. Our blockchain explorer guide explains how explorers use archive nodes to show you any transaction from any date.

Validator Nodes

On Proof of Stake chains such as Ethereum, a validator node stakes coins as a security deposit and is randomly chosen to propose and approve blocks. Ethereum requires 32 ETH to run a validator — a large sum, which is why many people stake through pools or services instead. Validators earn rewards for honest work and lose part of their stake (a penalty called slashing) if they cheat or go offline at the wrong time. Running a validator is a serious commitment: the machine must stay online almost all the time.

Mining Nodes

On Proof of Work chains such as Bitcoin, a mining node uses specialised hardware to compete for the right to create the next block. The competition is a race to solve a puzzle, and the winner earns the block reward plus transaction fees. Home mining of Bitcoin is no longer practical — the hardware is expensive and the electricity cost usually exceeds the reward — but mining still matters, because it is how Bitcoin’s nodes agree on the chain. Our mining guide covers the economics in detail.

Masternodes

Some chains, such as Dash, use masternodes: nodes that lock up a large amount of coins as collateral and provide extra services such as instant transactions or private transfers. Masternodes earn a share of the block rewards. They are less common than they were in 2018, but the idea — pay for extra service with locked collateral — still appears in newer chains under different names.

How to Run a Bitcoin Node at Home

Running a Bitcoin full node is the most practical first node project. It is well documented, the software is free, and it needs no stake and no special hardware. Here is the honest version of the process, including the parts tutorials usually skip.

Step 1: Get the Hardware

You need a computer that can stay on, a 1 TB SSD (the chain is roughly 700 GB in late 2026 and grows by about 50–60 GB a year), at least 8 GB of RAM, and a broadband connection with an unlimited or generous data allowance. An old laptop works, but a small dedicated machine such as a Raspberry Pi 4 or 5 with an external SSD is the popular choice because it uses very little electricity. Do not use a spinning hard drive: syncing the chain on one can take weeks instead of days.

Step 2: Install Bitcoin Core

Download Bitcoin Core from bitcoin.org — always from the official site, never from a link in a message. Install it, choose the data directory on your SSD, and start it. The first sync downloads the entire blockchain and checks every block from 2009 to today. This takes one to three days on a good connection with an SSD. Be patient: the software is verifying 17 years of history, one block at a time.

Step 3: Let It Sync and Stay Online

Once synced, your node uses very little bandwidth — a few gigabytes a month — and quietly checks each new block as it arrives. Open port 8333 on your router if you want your node to accept connections from other nodes and help the network; if you cannot, your node still works, it just helps fewer strangers. Leave the machine running. A node that is offline half the time still verifies your own transactions, but it contributes less to the network.

Step 4: Use Your Own Node

The real benefit comes when your wallet talks to your own node instead of someone else’s server. Wallets such as Sparrow and Electrum can connect to your node, which means your balance checks and transaction broadcasts are verified by your own machine. Nobody else learns which addresses are yours. That is the independence a node buys you: not money, but privacy and certainty.

How to Run an Ethereum Node

Ethereum is more demanding than Bitcoin, because an Ethereum node must run two pieces of software together: an execution client (such as Geth or Nethermind) and a consensus client (such as Lighthouse or Prysm). The execution client handles transactions and smart contracts; the consensus client handles agreement on the chain. Both must stay in sync with each other.

A full Ethereum node in late 2026 needs a 2 TB NVMe SSD, 16 GB of RAM, and a stable connection. The initial sync takes one to three days. Running a validator on top needs 32 ETH staked and near-perfect uptime, which is why most home users run a full node without validating. The Ethereum run-a-node guide is the official starting point, and it is kept up to date as the software changes. Our EVM guide explains the virtual machine your node runs with every block.

What Running a Node Costs — and What It Pays

The honest truth: an ordinary full node is a cost, not an income. You pay for hardware and electricity, and you earn nothing in coins. People run them for independence, privacy, and belief in the network — not for profit. Validators and miners are different: they earn rewards, but they also take on real costs and real risks, including slashing penalties and hardware failure. If someone promises you easy money from “running a node”, be suspicious. Real node income comes from staking or mining, both of which need serious capital and carry real risk.

Case Studies: Nodes in Real Life

Case Study 1: The Exchange That Caught a Fake Deposit

Problem: A small exchange received a deposit that looked valid in its wallet software but felt wrong — the amount did not match the customer’s claim. Strategy: The team checked the transaction against their own full node instead of trusting the wallet app. Execution: Their node showed the transaction had been broadcast but never confirmed, and the inputs had already been spent in a different transaction — a double-spend attempt. Results: The deposit was rejected before any funds were credited, saving the exchange the full amount. Lesson: A wallet app shows you what a server tells you. Your own node shows you what the chain actually says. For a business handling other people’s money, that difference is everything.

Case Study 2: The Student Who Ran a Node on a Raspberry Pi

Problem: A computer science student in Islamabad wanted to understand Bitcoin properly but could not afford a server. Strategy: He bought a Raspberry Pi 4 and a 1 TB SSD for under $150 and installed Bitcoin Core. Execution: The first sync took four days. He then connected his Sparrow wallet to the node and watched every new block arrive in real time. Results: He learned more about how Bitcoin works in one month of running a node than in a year of reading about it, and the project became the centrepiece of his job applications. Lesson: Running a node is the fastest way to learn a blockchain. Reading tells you the rules; a node shows you the rules being enforced.

Case Study 3: The Country That Tried to Block a Chain

Problem: A government blocked access to major crypto websites and pressured local exchanges to stop operating, hoping to cut its citizens off from a blockchain. Political pressure on crypto is nothing new — our history of Bitcoin after US midterms shows how elections and regulation have repeatedly shaken, but never stopped, the network. Strategy: Users switched to running their own nodes and connecting through them directly. Execution: Because nodes talk to each other peer to peer, with no central website to block, transactions kept flowing. Users broadcast through nodes in other countries, and the chain never noticed the ban. Results: The block failed. The chain kept working because no single point of failure existed to attack. Lesson: Decentralisation is not a slogan. It is a measurable property of a network: the number of independent nodes, spread across enough countries, that would all have to fail at once.

Case Study 4: The Staking Pool That Went Offline

Problem: A popular staking pool’s validators went offline during a data centre outage, and its users began losing small amounts to inactivity penalties. Strategy: The pool’s engineers had prepared for this: their validators ran in two separate locations with automatic failover. Execution: The backup site took over within minutes, and the pool’s downtime was limited to a single short window. Results: Users lost only a tiny amount in penalties, and the pool published a full incident report explaining what happened. Lesson: Running a validator is an operations job, not a set-and-forget investment. Uptime, monitoring, and backups are the real work — the staking itself is the easy part.

Tools for Running and Monitoring Nodes

Common Mistakes When Running a Node

Mistake 1: Using a slow hard drive. A spinning hard disk makes the first sync take weeks and the node struggle to keep up. Use an SSD. For Ethereum, use a fast NVMe SSD. This one choice decides whether the project is pleasant or miserable.

Mistake 2: Skipping backups of the wallet, not the chain. The blockchain itself needs no backup — you can always re-download it. What needs a backup is your wallet’s seed phrase and any validator keys. Lose those and the money is gone, however healthy your node is.

Mistake 3: Running a validator without monitoring. Validators must stay online and vote on time. If your machine reboots at 3 a.m. and nobody notices, you lose rewards and can face penalties. Set up alerts before you stake, not after your first missed duty. Our Ethereum ETF outflow report shows why validator economics and staking demand move with the wider market.

Mistake 4: Trusting a hosted node and calling it “your node”. Services such as Alchemy and Infura are convenient, but they are someone else’s node. They can see your requests, they can go down, and they can censor. They are fine for development. They are not the same as running your own.

Mistake 5: Forgetting the internet bill. A full node downloads hundreds of gigabytes during the first sync and a steady stream afterwards. On a capped mobile connection, that can be expensive. Use home broadband with a generous allowance, and check the sync size before you start. Our Bitcoin November outlook is a reminder that the network keeps growing whatever the price does — the chain only gets bigger.

Nodes, Decentralisation, and Why the Count Matters

People often ask how many nodes a chain has, as if the number alone proves decentralisation. It does not. What matters is who runs them, where they are, and how hard they are to control together. Ten thousand nodes all sitting in one company’s data centre are less decentralised than one thousand nodes spread across a hundred countries on home connections.

Bitcoin has tens of thousands of reachable full nodes, spread across the world, run by individuals, businesses, and hobbyists. Ethereum has thousands of full nodes and a smaller number of validators, with ongoing debate about how many run on the same few cloud providers. That debate is healthy: it is exactly the kind of question node counts are for. A chain whose nodes all depend on two cloud companies has a hidden point of failure, whatever its marketing says. If you want to understand why this matters for prices and for trust, our Galaxy bear market analysis and history of crypto cycles show how infrastructure strength decides which projects survive a downturn.

Frequently Asked Questions

What is a blockchain node in simple terms?

A node is a computer that runs a blockchain’s software and keeps a copy of its records. It checks every transaction against the rules and shares the results with other nodes. Together, the nodes are the blockchain.

Do I need a node to use crypto?

No. Your wallet app uses someone else’s node — usually run by the wallet company or a service such as Infura. Running your own node is optional, but it gives you privacy and lets you verify transactions yourself instead of trusting a third party.

Can I earn money by running a node?

An ordinary full node earns nothing. Validators earn staking rewards, and miners earn block rewards, but both need significant capital and carry real risk. Anyone promising easy money from “running a node” is usually selling something.

How much storage does a Bitcoin node need?

About 700 GB in late 2026, growing by roughly 50–60 GB a year. Use a 1 TB SSD at minimum. An Ethereum full node needs a 2 TB NVMe SSD.

How long does it take to sync a node?

One to three days for Bitcoin or Ethereum on a good connection with an SSD. On a slow hard drive or a weak connection it can take a week or more.

What is the difference between a full node and a light node?

A full node stores the entire chain and verifies everything itself. A light node stores only block headers and asks full nodes for details. Full nodes are independent; light nodes are fast and fit on a phone.

Can I run a node on a Raspberry Pi?

Yes, for Bitcoin — a Raspberry Pi 4 or 5 with a 1 TB SSD works well and uses very little electricity. Ethereum needs a more powerful machine, and Solana validation needs a server.

What happens if my node goes offline?

Nothing bad happens to the network — the other nodes carry on. Your own node simply falls behind and must catch up when it returns. Only validators face penalties for going offline.

Is running a node legal?

In most countries, yes. Running a node is just running software and storing public data. Rules vary by country, so check your local position, especially if you plan to validate or mine commercially.

How many nodes does Bitcoin have?

The number changes daily, but there are tens of thousands of reachable full nodes worldwide, plus many more behind firewalls that are not counted. The exact figure matters less than the spread: they run in dozens of countries, on many kinds of hardware, owned by many different people.

Conclusion

A blockchain node is the humble worker that makes everything else possible. It stores the records, checks the rules, and talks to its neighbours — and because thousands of them do it independently, no company, government, or hacker can rewrite the history they hold together. You do not need to run one to use crypto, but running one changes how you understand it. The chain stops being an abstract idea and becomes a machine on your desk, humming quietly, verifying every block as it arrives.

Start small. A Bitcoin node on a mini PC costs less than a mid-range phone, takes a weekend to set up, and teaches more about how crypto really works than a year of price charts. And once your own node has verified your first transaction, you will never again have to take anyone’s word for where your money is. Which chain will you help keep honest first?