Nimiq vs Bitcoin and Ethereum: what actually differs

The comparison, in one table

Every Nimiq figure below comes from the node software's own policy module, and every Bitcoin and Ethereum figure from that project's own documentation. Nothing here is taken from a third party comparison page, which is where most of the wrong answers about Nimiq come from.

The rest of this page is that table with the reasoning put back in, because a row of a table tells you what differs and not why anybody chose it.

First, a correction: Albatross is not a hybrid

Before any comparison is worth reading, one claim has to go. A striking number of articles about Nimiq, including several that rank well, state that Albatross combines proof of work and proof of stake. It does not, and this is not a matter of interpretation.

Nimiq launched as a proof of work chain in April 2018 and ran that way for six years. On 19 November 2024 it migrated to Albatross at proof of work block 3,456,000. That height is not folklore: it is written into the node software as genesis_block_number in the mainnet policy, and it is the number every block height on the current chain is counted from. There is no residual mining, no hashing component, and no hybrid mode. The proof of work chain ended and a proof of stake chain continued from its final state.

Two things probably keep the error alive. Nimiq spent most of its written history as a proof of work chain, so a lot of the corpus describes a network that no longer exists. And Albatross genuinely does combine two things, just not those two: it combines Byzantine fault tolerant voting with a fast single producer block pipeline. Somewhere in the retelling, "combines two mechanisms" became "combines proof of work and proof of stake".

Consensus: energy, capital, and who gets asked

All three chains answer the same question, which is who gets to append the next block.

• Bitcoin asks the machine that did the most work. Miners race to find a hash below a target, and the difficulty adjusts so that the race takes about ten minutes. Security comes from the cost of redoing that work, which is why the Bitcoin developer guide describes the cost of altering a block as rising with every block stacked on top of it.
• Ethereum asks a validator chosen by the protocol. Time is cut into 12 second slots and 32 slot epochs, and a validator activated with 32 ETH proposes and attests. Security comes from the stake that misbehaviour destroys.
• Nimiq asks whichever validator holds the current slot. There are 512 slots, shared out among validators in proportion to stake at each epoch boundary, and the slot owner list is reshuffled for every single micro block. Security also comes from stake, but the block pipeline is much tighter.

So Nimiq and Ethereum are in the same family and Bitcoin is not. That family resemblance is real and worth saying, because the interesting differences between Nimiq and Ethereum are not about consensus at all.

UTXO versus accounts, and why it shows up in a wallet

Bitcoin does not store balances. It stores unspent transaction outputs, and your balance is whatever the wallet finds when it adds up the outputs you can spend. Spending means consuming whole outputs and creating new ones, including change back to yourself. It is a good model for auditing and for privacy techniques, and it is the reason Bitcoin fee estimation depends on how many outputs your wallet has to gather.

Nimiq and Ethereum both store an account with a balance, and a transfer subtracts from one and adds to the other. No change outputs, no coin selection. For a payments chain that is the simpler model, and it is why a Nimiq address behaves the way most people already expect an account to behave.

Where Nimiq then parts company with Ethereum is what else an account can be. Ethereum accounts can hold arbitrary programs. Nimiq supports exactly four account types: a basic account, a hashed time locked contract, a vesting contract, and the staking contract. There is no virtual machine for general programs. That is a deliberate limit, and it cuts both ways: a much smaller protocol to audit and attack, and no ability to build the sort of thing Ethereum exists for.

Where a node can run, which is the actual differentiator

If you take one row from the table, take this one. It is the design decision Nimiq is organised around, and neither of the others has an equivalent.

Nimiq's client is compiled to WebAssembly and runs as a light or pico node inside an ordinary browser tab. It peers with the network directly, syncs enough of the chain to check the current tip for itself, and broadcasts transactions without a server in the middle. Opening a Nimiq wallet is opening a web page, and that page is a network participant rather than a front end for somebody's API.

Neither Bitcoin nor Ethereum works that way in practice. A Bitcoin full node is a program you install and a chain you download. Ethereum has browser oriented light client work in progress, but the normal way a web page reads Ethereum is to send JSON-RPC to a provider and believe the response. That is a trust relationship, and usually an undisclosed one.

Two honest caveats, because this row is the one most easily oversold:

• A light node is not an archive node. It holds headers, not the full history, so deep historical lookups still come from an indexing service. Our own block explorer is exactly such a service.
• Plenty of Nimiq apps use an API anyway, this site included, because it is faster and cheaper. The point is that the browser client makes that a choice rather than the only option.

Finality: three genuinely different promises

"Confirmed" means something different on each of these chains, and conflating them is how people end up surprised.

Bitcoin is probabilistic and never finishes. A transaction becomes harder to reverse as blocks pile on top, and the whitepaper works out the probability directly: it falls exponentially with each confirmation. It never reaches zero. The familiar six confirmations, roughly an hour, is a convention people settled on, not a protocol rule.

Ethereum votes, and takes a while. Validators attest to checkpoints and a block finalises once the required supermajority has voted across two epochs. Ethereum's own roadmap page states it plainly: it takes about 15 minutes for a block to finalise, and removing that wait is what the single slot finality work is for.

Nimiq votes too, but every batch. Sixty blocks make a batch, one per second, and the last block of each batch is a macro block confirmed by at least two thirds of the 512 validator slots. Once that vote lands the whole batch is final, and the protocol does not reverse it. So the wait is about a minute, and at the end of it you have a voted guarantee rather than a probability.

Which is genuinely better depends on what you are doing, and neither of the fast answers is free. Voted finality needs a known validator set and the assumption that no more than a third of it is malicious. Bitcoin needs no such assumption, which is the trade it makes for the hour. How Albatross works takes the Nimiq side of this apart properly.

What it costs to take part

Comparing the price of entry is where the three designs feel least alike, and no figure here is a price in money.

The right hand column is the one that differs most and gets discussed least. Nimiq's delegation is part of the protocol itself: the staking contract records that your address backs a particular validator, your NIM never leaves your account, and you undo it with protocol transactions rather than by asking anyone. Bitcoin has no such concept. On Ethereum, delegating without 32 ETH generally means a pool or a liquid staking token, which is a product built on top rather than a protocol feature.

Both Nimiq deposit figures are constants in the node software, VALIDATOR_DEPOSIT and MINIMUM_STAKE, stored in Luna and worth 100,000 NIM and 100 NIM respectively. How to stake NIM covers the delegation path, and our validators page lists who you could delegate to.

Energy, stated factually

This is usually where a comparison page starts shouting. We are going to describe the architecture and decline to print a multiplier.

The architectural difference is not in dispute. Bitcoin's security budget is energy: the protocol is designed so that rewriting history means redoing computational work, and the difficulty adjustment guarantees that miners collectively burn enough electricity to keep block times near ten minutes however much hardware joins. That is the mechanism working as intended, not a bug in it.

Proof of stake replaces that cost with capital at risk. A Nimiq validator is a server that needs to stay online and answer on time; the thing that would be destroyed by misbehaviour is stake and forfeited rewards, not fuel. Ethereum works the same way. So the energy question between these three is really the proof of work question, and it has one answer, not three.

What we will not do is tell you Nimiq uses some specific fraction of Bitcoin's energy. Every credible estimate of Bitcoin's consumption is a modelled range rather than a measurement, the published comparisons differ by large factors, and we have measured neither. An honest architectural claim is worth more than a number we cannot stand behind.

What Nimiq gives up

A comparison page written by people who run a validator on one of the three chains should say where their own chain is weaker, so here it is.

• Far less economic weight securing it. Bitcoin and Ethereum are secured by vastly larger amounts of capital and hardware. A smaller network is a cheaper network to attack, and no amount of protocol elegance changes that arithmetic.
• A small validator set. Nimiq's registered validators number in the dozens, not the thousands. Our staking in numbers case study measures exactly how concentrated it is, and it is more concentrated than anyone would like.
• No general programmability. If you want to build something that is not a payment, Nimiq is the wrong chain and Ethereum is probably the right one.
• A much smaller ecosystem. Fewer independent client implementations, fewer tools, fewer people reading the code. This hub exists partly because so little accurate writing about Nimiq exists at all.

Against that, the thing Nimiq actually has: an account you can create in a browser in under a minute with no download and no identity check, transactions final in about a minute, and self custodial staking built into the protocol. Whether that trade is worth making is your call and not something this page will make for you.

Check it yourself

None of the Nimiq claims above need to be taken on trust. Open our block explorer and watch a block arrive roughly every second, then watch the height cross a multiple of 60 and see the macro block that closes the batch. The validators page shows the real validator set and what each one holds. Every protocol constant quoted here is a named line in a public source file, linked in the sources below.

If you arrived without knowing what Nimiq is, start there. For the consensus mechanism in detail, read how Albatross works. For the supply cap in the table, NIM tokenomics explains how a cap that is never actually reached still functions as a cap. The rest is in the Learn hub.

We publish no price figures anywhere on this site, nothing here is a recommendation to hold any of these three coins, and nothing on this page is financial advice.