Current Bitcoin difficulty, a chart of every retarget and a forecast of the next one. Data comes from RuPool's own full nodes and refreshes every minute.
134.82T
Exact value: 134,823,837,323,273
1974 of 2016 blocks left
2.1%
Estimated Nov 28, 2026, 04:09 UTC
Average block time this epoch
—
target 10:00Network hashrate (estimate)
964.97 EH/s
Previous retarget
-9.13%
Nov 14, 2026Next retarget block
#977,760
| Date (UTC) | Block | Difficulty | Change |
|---|---|---|---|
| Nov 14, 2026 | 975,744 | 134.82T | -9.13% |
| Oct 29, 2026 | 973,728 | 148.36T | +5.54% |
| Oct 16, 2026 | 971,712 | 140.58T | +5.92% |
| Oct 3, 2026 | 969,696 | 132.72T | -0.03% |
| Sep 19, 2026 | 967,680 | 132.76T | +4.16% |
| Sep 5, 2026 | 965,664 | 127.45T | +1.31% |
| Aug 23, 2026 | 963,648 | 125.81T | -1.31% |
| Aug 8, 2026 | 961,632 | 127.48T | +0.99% |
| Jul 25, 2026 | 959,616 | 126.23T | -0.74% |
| Jul 11, 2026 | 957,600 | 127.17T | -5.00% |
| Jun 27, 2026 | 955,584 | 133.87T | +7.15% |
| Jun 14, 2026 | 953,568 | 124.93T | -10.09% |
Bitcoin difficulty tells you how many times harder it is to find a block now than at the very start: the network launched in 2009 at difficulty 1. A block is valid when its SHA-256 hash is below the target, and difficulty is the ratio of that original target to the current one.
The number has a handy physical meaning: to find one block, the network computes on average difficulty times 2 to the power of 32 hashes. Divide that by the 600-second target block time and you get an estimate of network hashrate. That is why the BTC difficulty chart is essentially a smoothed chart of the combined power of every Bitcoin miner.
Bitcoin adjusts difficulty at every block whose height is divisible by 2016. The node takes the actual length of the epoch that just ended and compares it with the planned 14 days: if blocks came faster, difficulty rises proportionally; if slower, it falls. To keep a glitch from crashing or spiking the network, a single adjustment is capped at a factor of four either way.
Bitcoin Core has a long-known quirk: the epoch is measured from its first block to its last, which is 2015 intervals, yet compared with the planned time for 2016. Because of this off-by-one, difficulty ends up about 0.05% higher than ten-minute blocks would require. Nobody changed a consensus rule over such a small effect, and our forecast calculates exactly the way the network does.
| Parameter | Bitcoin |
|---|---|
| Algorithm | SHA-256 |
| Target block time | 10 minutes |
| Epoch length | 2016 blocks, about 14 days |
| Maximum change per adjustment | −75% to +300% |
| Genesis block difficulty | 1 |
While an epoch is running, the next Bitcoin difficulty is already partly decided: we know how long its blocks have taken. The forecast at the top of the page takes the time from the epoch's first block to the latest one, divides it by the number of intervals and compares the average with 10 minutes. Below that means an increase is expected, above it a decrease. The estimated date is the remaining blocks multiplied by the same average.
The start of an epoch is too random, so the forecast appears after 144 blocks, roughly a day after the adjustment. It sharpens with every block but stays an estimate: if large farms come online or go down for maintenance in the last days, the result will move. The data comes from our own Bitcoin full node and refreshes every minute.
The chart shows every adjustment since the network launched. Almost all of it is growth: each ASIC generation is faster and more efficient, and new capacity arrives faster than old machines retire. Plateaus and dips tend to line up with falling prices or rising power costs, when the least efficient hardware stops paying for itself.
The sharpest single drop came in July 2021: −27.94% after mining was banned in China. The hashrate didn't vanish; it was moving with the hardware to other countries, and within months difficulty was climbing again. The table of the last 12 adjustments covers roughly the past six months.
At a constant hashrate, an ASIC finds a share of blocks inversely proportional to difficulty. A 5% increase cuts bitcoin income by about 4.8%, and a 5% decrease raises it by about 5.3%, assuming the same price and network fees. That is why payback on new hardware is best estimated with room for difficulty growth rather than at today's value.
RuPool pays with the FPPS scheme and a 1% fee: the reward for each accepted share is calculated from network difficulty, so income per terahash changes after every adjustment. The profitability calculator shows what a specific ASIC model earns.
At the next block whose height is divisible by 2016. The blocks remaining and the estimated date are shown at the top of the page. The date is based on the current epoch's average block time, so it can shift by hours either way.
Over the first few dozen blocks the average is too random: a handful of fast blocks in a row would suggest an increase of tens of percent. After 144 blocks, about a day, the estimate becomes meaningful.
Consensus rules cap a single adjustment at a factor of four: no more than +300% and no less than −75%. In practice, recent adjustments usually stay within a few percent, and moves above 10% are rare.
−27.94% in July 2021, after mining was banned in China. It is the largest single-adjustment decrease in the network's history: a big part of the hashrate was offline for months while hardware was being relocated.
Difficulty rises when network hashrate grows and blocks start arriving faster than every 10 minutes. Every new ASIC generation delivers more terahashes per watt, and as long as mining stays profitable, capacity keeps being added.
From the Bitcoin full node RuPool runs for its own operation. Values refresh every minute, and no third-party APIs are used.