Bitcoin Mining Explained: How New Bitcoin Are Created

Short answer: In bitcoin mining, miners bundle new transfers into a block and search, by trial and error, for a number that makes the block valid. Whoever finds it first receives 3.125 new bitcoin plus fees (since 2024). The computing work protects the network against forgery, and the difficulty adjustment keeps the pace at one block every 10 minutes.

You may have heard that bitcoin "eats electricity", that miners "solve puzzles" or that you can make money with a machine in your basement. Most of that is half right. This article explains bitcoin mining without jargon: what a miner really does, who takes part in the network, what a block earns today, why this work protects your bitcoin and why mining at home in Germany and Switzerland almost never pays.

What is bitcoin mining?

Bitcoin has no bank that books transfers. Yet the whole network has to agree on who sent how much to whom, and in what order. That is exactly what bitcoin mining does: miners collect the waiting transfers, check them and write them as a new block onto the end of the blockchain.

So that not just anyone can do this at will, a block costs real work. A miner has to compute billions of attempts before it can present a valid block. Whoever manages it may credit itself a reward. That is how all new bitcoin are created, and the only way.

The name comes from gold mining, a comparison made in the white paper: like gold, bitcoin is mined with effort, and what is left in the ground keeps getting smaller.

Rule of thumb: Miners are bitcoin's bookkeepers. They are paid with new bitcoin and fees, and their work makes the books forgery-proof.

If terms like blockchain or wallet are still new to you, start with the introduction What is bitcoin? or the basics in the academy.

How bitcoin mining works, step by step

A block is created in five steps. They run at every miner at the same time, day and night:

1. Collect transfers. Every new bitcoin transfer first lands in a waiting area, the mempool. From it the miner picks the transfers with the highest fees per byte until the block is full.
2. Check the rules. For each transfer it checks that the coins exist, have not yet been spent and that the signature is correct. One invalid transfer would make the whole block invalid.
3. Enter the reward. As the first line, the miner credits itself the reward: the new bitcoin plus all the block's fees.
4. Roll the dice. Now the real work begins. From the block header the miner computes a fingerprint, the hash. If the hash is below a target value, the block is valid. It almost always lies above it. Then the miner changes a number in the block header, the nonce, and computes again.
5. Broadcast the block. If it finds a fitting hash, it sends the block to the network. Every node checks within a fraction of a second that everything is correct and appends the block. All miners start on the next block.

Nobody can predict the hash: change one character in the block and a completely different fingerprint comes out. There is no clever way to the goal, only trial and error, more a game of dice than a puzzle.

What miners earn: block subsidy and fees

The reward for a block has two parts. The block subsidy is newly created bitcoin, fixed in the code. The transaction fees are paid by the users whose transfers are in the block.

Since the halving in April 2024, the subsidy has been 3.125 bitcoin per block. Fees are small at the moment: in the 144 blocks up to Nov 12, 2026, roughly one day, they averaged 0.025 bitcoin per block. That is less than 1% of the reward.

Block data: mempool.space, blocks 975’469 to 975’612. Price: €64’542 and CHF 59’645 per bitcoin (mempool.space, Nov 12, 2026).

All the miners in the world compete for this roughly €29 million a day. Whoever has cheaper electricity and more efficient machines keeps more of it.

What one unit of computing power earns, miners call the hashprice. On Nov 12, 2026, 1 TH/s, that is one trillion attempts per second, earned about 48 satoshi a day, just over 3 cents.

The halving: why the reward keeps shrinking

Every 210’000 blocks, roughly every four years, the subsidy halves. That is how Satoshi Nakamoto limited the supply to just under 21 million bitcoin. No miner and no developer can change this, because every node would reject a block with too high a reward. All the dates, and what happened after past halvings, are in the bitcoin halving at a glance.

Up to block 975’612 (Nov 12, 2026), about 20.11 million bitcoin have been created, just under 96% of all that can ever exist. The remaining roughly 0.89 million come in ever smaller portions over more than a hundred years. Around the year 2140 the subsidy reaches zero.

For miners, every halving means the largest part of their income halves overnight. Older machines then often run at a loss and are switched off. In the long run, fees are supposed to replace the shrinking subsidy. Whether that will be enough is one of bitcoin's open questions.

Why this fixed issuance matters for your purchasing power is shown in the article Inflation and loss of purchasing power, compared with the euro and franc.

Difficulty and hashrate: the 10-minute beat

The hashrate is the combined computing power of all miners. Mempool.space estimated it at about 936 EH/s on Nov 12, 2026, that is 936 quintillion attempts per second. A year earlier the daily average was about 1’090 EH/s, most recently about 980 EH/s. Since its record high in October 2025, bitcoin has lost about 40%, and some miners have switched off.

If nothing counteracted it, blocks would come faster with more miners and more slowly with fewer. That is what the difficulty adjustment is for: every 2’016 blocks, each node calculates whether blocks came faster or slower than every 10 minutes on average, and makes the dice-rolling harder or easier accordingly. So the beat stays the same, however many miners take part.

How the adjustment calculates exactly, and which records there have been, is explained in the article Difficulty adjustment: the heart of bitcoin.

Why bitcoin mining protects the network

The computing work has a purpose beyond issuing new coins: it makes the past expensive to forge. Satoshi Nakamoto called the method proof of work.

Suppose someone wants to reverse a payment they sent you three blocks ago. They would have to recompute that block and all three following ones, and then overtake the honest miners who keep building on the real chain. The network always follows the chain with the most work. Without more than half of the world's computing power, according to the calculations in the white paper, this practically never succeeds.

At 936 EH/s that would mean millions of specialized machines and electricity on the scale of a mid-sized country, just to reverse your own payments. And even then the attacker could not steal anyone else's bitcoin, because they lack the keys. How to protect your own keys is in the guide Storing bitcoin safely.

Important: miners do not set the rules. Each node decides for itself whether a block is valid. A miner that credits itself more bitcoin or includes invalid transfers only wastes its electricity, because its block is rejected.

Who mines bitcoin today

Mining used to run on ordinary computers. Satoshi Nakamoto found the first blocks in 2009 with a processor. From 2010, graphics cards computed faster; since 2013, ASICs have dominated: machines that can do only one thing, compute SHA-256 hashes, but do it millions of times more efficiently than a PC.

Today, mining is done mostly by companies in large halls, often where there is surplus electricity: at hydropower plants, wind and solar farms, gas fields or in countries with cheap tariffs. Many are listed on US stock exchanges. After China's mining ban in 2021, computing power spread widely to the US, Russia, Kazakhstan and other countries.

Almost all miners join a pool. It bundles the computing power of many participants and divides the reward by share. That way a single miner gets small, regular payments instead of a tiny chance at a whole block.

887 blocks in the 7 days to Nov 12, 2026, mempool.space.

The concentration on a few pools is rightly debated. But a pool is not a data center, it is an association of many miners. They can switch to another pool within minutes if theirs misbehaves.

Energy use: how much electricity does bitcoin mining need?

Mining needs electricity, and a lot of it. That is not a flaw but the core of the protection: something that would be cheap to forge would offer no security. The question is how much and from where.

From the hashrate and the machines' efficiency you can roughly estimate consumption. At 936 EH/s and a fleet average of 15 to 25 joules per TH, that comes to about 120 to 200 terawatt hours a year. That is roughly a quarter to just under half of Germany's electricity consumption. The University of Cambridge's Cambridge Bitcoin Electricity Consumption Index (CBECI) publishes an ongoing estimate in this range.

Unlike a data center, a miner doesn't have to be in a particular place and can switch off at once. That is why miners deliberately seek electricity nobody else wants: surplus wind power at night, curtailed solar peaks or gas that would otherwise be flared. Where the electricity actually comes from, and which headlines get the math wrong, is shown in the article Bitcoin's energy consumption.

Bitcoin mining at home: is it worth it?

The honest answer for Germany, Austria and Switzerland: almost never, if you pay normally for your electricity. The reason is simple: professional miners often pay 3 to 6 cents per kWh, households many times that.

There are exceptions. If you would otherwise feed your own summer solar power into the grid for a few cents, or if you heat anyway in winter and use the miner's waste heat, the math is different. Even then: an ASIC is as loud as a vacuum cleaner and produces a lot of heat. And every difficulty increase and every halving shrinks the income further.

What solar surplus, waste heat and taxes change in this calculation is worked through in the article Bitcoin mining at home in Germany and Switzerland.

Do the math yourself with your machine, your electricity price and your time frame. The calculator also shows whether simply buying would have brought more bitcoin:

Solo mining: the lottery

Instead of in a pool, you can also mine alone. Then you either get the whole block, about 3.15 bitcoin, or nothing at all. With 200 TH/s the chance of a block is about 1.1% a year. On average you would have to wait about 90 years. How big your chance is with your machine is shown by the solo mining calculator. What a hit brings, how often solo miners really win and how to get started is explained in the article Solo mining: the bitcoin lottery.

What mining means for you as a bitcoin holder

You don't have to mine to benefit from mining. For you as a long-term holder it does three things: it protects your coins against later forgery, it issues new bitcoin on a fixed schedule, and it keeps the network running without a boss.

The hashrate is not a buy signal. Miners react to the price with a delay: when prices rise they connect new machines, when they fall they switch off old ones. Where the price stands relative to its long-term trend is shown by the Power Law analysis with floor and zones; how we calculate is in the methodology. The model describes the past and is no guarantee.

How bitcoin achieves fast and cheap everyday payments despite limited blocks is explained in the article Lightning Network explained. Why it is worth checking the rules yourself with your own bitcoin node is shown in a separate article. How these rules are extended through a soft fork such as Taproot is explained in another. Why quantum computers could threaten signatures but not mining is explained in Quantum computers and bitcoin. More articles on energy and fees are collected in the category Mining & technology.