Bitcoin mining difficulty explained: why coin output shrinks
Difficulty is the bar every Bitcoin block has to clear: here is how it resets about every two weeks, how to read it on mempool.space, and why it usually shrinks what a fixed amount of hashrate mines.
Published by Minevana · October 14, 2026 · 8 min read
The short answer
Mining difficulty is a number that sets how many hashes, on average, it takes to find a Bitcoin block. Every 2,016 blocks (about two weeks) the network resets it so blocks keep arriving roughly every 10 minutes. When more machines join, difficulty rises, so the same hashrate wins a smaller share of new bitcoin. That is why a fixed miner's coin output usually shrinks over time.
What is Bitcoin mining difficulty?
To add a block, a miner has to find a block header whose hash comes out below a target number. There is no shortcut: machines make trillions of guesses a second until one lands under the target. Difficulty is that target expressed the other way round. A higher difficulty means a lower target, and more guesses needed on average.
The numbers are large. Each unit of difficulty stands for about 4.3 billion (2³²) hashes per block on average. As of 14 October 2026, difficulty is about 132.72 trillion (132.72 T), set at block 969,696 on 3 October 2026. Multiply that by 2³² and divide by the 600 seconds in ten minutes, and you get the network hashrate that difficulty implies at a 10-minute pace: about 950 [EH/s](/glossary#eh-s), or 950 million TH/s.
Nobody sets this number by hand. There is no company and no vote: every Bitcoin node applies the same rule to the same block timestamps, so they all arrive at the same new difficulty.
How does the difficulty adjustment work?
- 1
Count 2,016 blocks
The adjustment happens once every 2,016 blocks: 14 days of 144 blocks a day at the 10-minute target. As of 14 October 2026, the current period began at block 969,696, so the next adjustment falls at block 971,712.
- 2
Measure how long they took
Nodes compare the time the period actually took, read from block timestamps, with the target of 1,209,600 seconds: exactly two weeks.
- 3
Scale difficulty by the gap
If the period took 12.6 days, difficulty is multiplied by 14 ÷ 12.6, about 11% higher. If it took 15.6 days, difficulty drops to 14 ÷ 15.6 of its old level, about 10% lower. That is close to what happened in June 2026, when a period ran about 15.6 days and difficulty fell 10.09%.
- 4
Never more than fourfold in one step
The rule caps each adjustment at four times higher or a quarter of the old level, so one odd period cannot swing difficulty without limit. Real adjustments are usually far smaller: the largest single drop on record was about 28%, in July 2021.
How has Bitcoin mining difficulty changed over time?
Difficulty at a few points in Bitcoin's history, read from block data and rounded:
- January 2009 (block 0): 1, the starting value.
- July 2016 halving (block 420,000): about 213 billion.
- May 2020 halving (block 630,000): about 16.1 trillion.
- April 2024 halving (block 840,000): about 86.4 trillion.
- 29 October 2025 (block 921,312): about 156 trillion, the highest on record as of October 2026.
- 3 October 2026 (block 969,696): about 132.7 trillion. That is about 15% below the peak and about 10% below where 2026 began, but still about 54% above the April 2024 level.
Why does mining difficulty usually rise?
Bitcoin's whitepaper says the moving difficulty target is there to make up for increasing hardware speed and for varying interest in running nodes, which in the whitepaper means mining, over time. Both have mostly pushed the same way. Each new generation of mining machines (ASICs) does more hashes for the same electricity, and when the bitcoin price makes mining pay, operators plug more of them in. More hashrate means faster blocks, and the next adjustment pushes difficulty up.
Difficulty does fall. When the price drops or power gets expensive, the least efficient machines are switched off, blocks slow down and the next adjustment eases. In July 2021, China's mining crackdown took so much hashrate offline that difficulty dropped about 28% in a single step. In June 2026, what Galaxy Research called a price-driven margin squeeze cut it by 10.09%. In mid-October 2026, difficulty was about 12% lower than on the same dates a year earlier (150.84 T).
So far, falls like these have been interruptions in a long climb rather than reversals of it. Nothing says that pattern has to continue, so treat any forecast of future difficulty as an assumption, not a fact.
What moves mining income, and who controls it?
| What changes | How often | Who or what controls it | Effect on coins from a fixed hashrate |
|---|---|---|---|
| Network hashrate | All the time | Every miner in the world, switching machines on and off | Indirect: it drives the next difficulty change |
| Difficulty | Every 2,016 blocks (about two weeks) | A fixed rule in every node's software | Higher difficulty, fewer coins |
| Block subsidy | Halves every 210,000 blocks (about four years); next at block 1,050,000 | The Bitcoin protocol | A halving cuts the new-coin part in half |
| Transaction fees | Every block | Users bidding for space in blocks | Higher fees add a little to each block's reward |
| Bitcoin price | All the time | The market | No direct effect on coins mined; changes what they are worth |
| Your hashrate | Fixed when you buy the machine or plan | You, at the time of purchase | Stays the same; it does not grow on its own |
Only the last row is in your hands, and only when you buy.
How to read a difficulty chart on mempool.space
- 1
Open the hashrate and difficulty chart
Go to mempool.space/graphs/mining/hashrate-difficulty and pick a range: 1M, 3M, 6M, 1Y, 2Y, 3Y or ALL. Short ranges show the recent steps; ALL shows the long-run climb.
- 2
Tell the lines apart
The Difficulty line is flat between adjustments and jumps up or down at each one. Hashrate is a thin, jagged estimate worked out from how fast blocks are being found, and Hashrate (MA) is a smoothed average of it. Hover over the chart to see both values for a given day.
- 3
Watch hashrate lead and difficulty follow
Hashrate moves first; difficulty catches up at the next step. If the smoothed hashrate line has climbed since the last step, blocks are coming faster than every 10 minutes and the next step will probably be up. If it has slid, the next step will probably be down. Ignore single-day spikes: the raw estimate is noisy.
- 4
Check the live estimate
On mempool.space/mining, the Difficulty Adjustment panel shows how many blocks remain in the period, the estimated change and the previous change, and an Adjustments table lists recent adjustments with block height, how long ago each happened, the new difficulty and the percentage change. The estimate shifts with every block and is only final at the adjustment block.
What does difficulty mean for a fixed amount of hashrate?
A fixed hashrate's coin output moves opposite to difficulty: if difficulty doubles, the same machine mines half as many coins. A real, dated example: at the 9 October 2024 difficulty (92.05 T), 1 TH/s worked out to about 68 sats of block subsidy per day on average. At the 3 October 2026 difficulty (132.72 T), the same 1 TH/s works out to about 47 sats. That is about 31% fewer coins, with nothing about the machine changing.
Those figures are network arithmetic (1 sat is 0.00000001 BTC). They leave out transaction fees and any pool or operator costs, they are not what any particular plan pays, and they change every two weeks.
A second force arrives on a schedule. The next halving, at block 1,050,000, is expected around April 2028; the exact date depends on block times. From that block the subsidy drops from 3.125 to 1.5625 BTC, so the same share of the network mines half the new coins. Any 24-month mining contract that starts in late 2026 is likely to run through it. Difficulty is one of the six variables covered in is cloud mining profitable in 2026?
If difficulty keeps changing for 24 months
How a fixed hashrate's coin output would compare under steady monthly difficulty changes:
| If difficulty each month… | Coins per day after 24 months, vs at the start | Coins over all 24 months, vs flat difficulty |
|---|---|---|
| Falls 1% | 127% | 113% |
| Stays flat | 100% | 100% |
| Rises 1% | 79% | 89% |
| Rises 2% | 62% | 80% |
| Rises 3% | 49% | 72% |
Illustration only: block subsidy held constant, transaction fees and the halving ignored. For scale, difficulty rose about 44% from October 2024 to October 2026 (about 1.5% a month on average), even though it fell about 12% over the last twelve of those months.