Bitcoin’s 20 Millionth Coin Has Been Mined – Only 1 Million BTC Left to Ever Exist

What does “20 million BTC mined” actually mean?

When people hear “20 million BTC mined,” many think Bitcoin is almost finished. That is partly true, but not in the way most people mean it. The network is not close to shutting down. Mining will keep going for decades. What this number really tells us is that Bitcoin’s issuance era is now far along. Kraken said more than 95% of all coins that will ever exist already exist. TradingView put the figure at 95.24% at the milestone block. Bitbo’s live network page showed about 95.25% issued on March 17, 2026.

That matters because Bitcoin is built around a fixed upper limit. No board can vote to print more. No central party can decide to loosen supply because demand rises. The code keeps issuing smaller and smaller amounts at a known pace. Binance’s March 2026 write-up said only 1 million units remain to be issued by around 2140, and that today’s block reward is just 3.125 BTC.

There is another layer here too. “Mined supply” and “spendable supply” are not the same thing. TradingView cited estimates from River Financial and Chainalysis saying that somewhere between 2.3 million and 3.7 million BTC may already be gone for good because people lost keys, forgot passwords, or never passed on wallet access. That means the practical float may be far smaller than the raw supply figure suggests.

A simple way to look at the milestone is this:

• Bitcoin has not reached its full cap yet
• fresh supply is still coming out
• fresh supply is now small compared with total supply
• truly usable supply is likely lower than 20 million

The shift becomes easier to see in a basic comparison. Historical supply steps below come from milestone reporting and Bitcoin’s fixed issuance schedule.

So the real meaning of this event is not “Bitcoin is nearly over.” It is “Bitcoin’s supply growth is already very small, and it will keep getting smaller.” That is the part with the biggest long-run impact.

How Bitcoin mining works 

The process of Bitcoin mining becomes easier to understand after its complex components have been simplified. People send transactions across the network. The transactions remain in a waiting state until they get combined with other transactions into a block. The mining machines compete to locate a valid hash that will secure the block.

The network accepts the first valid block which gets added to the blockchain and gives a reward. Bitcoin.org describes mining as the system that verifies transactions through distributed consensus while maintaining blockchain order and enabling computer systems to share a common database.

The process works like this:

1. users send BTC transactions
2. the transactions spread across the network
3. miners bundle many of them into a candidate block
4. machines keep trying hashes until one fits the network rule
5. the winning block is checked by the network
6. the block is added, and the winning miner gets the reward

The reward has two parts. One part is the fixed subsidy, which is 3.125 BTC right now. The other part is the transaction fees inside that block. Bitinfocharts showed that on March 17, 2026, the average reward per block was about 3.125 plus 0.0221 BTC in fees, with fees making up about 0.65% of the total block reward over the prior 24 hours.

That also explains why Bitcoin miners are better viewed as security providers than coin printers. They spend real money on ASICs, cooling, buildings, staff, and power so the chain stays hard to attack and hard to rewrite. Bitcoin.org says mining also acts like a competitive lottery that keeps any one party from easily adding blocks one after another or rolling back its own spending.

The network keeps the rhythm close to one block every ten minutes through difficulty adjustment. Lightspark says difficulty changes every 2,016 blocks, about every two weeks, so block production stays near target even if a lot of hash power joins or leaves. That is why the issuance schedule can stay steady even while the mining business around it changes all the time.

As of March 17, 2026, Bitbo showed difficulty at about 145.04 trillion, 24-hour hash rate near 859 EH/s, and hash price around $3.11 per TH per day. That tells you how industrial this has become. A single modern machine can still produce income, but it is competing against one of the most powerful computing networks ever built.

Why only 1 million BTC is left and why it will take decades

The statement that there exists a remaining supply of 1 million BTC presents an urgent and intense sound, yet the cryptocurrency Bitcoin has an extended period before its final coins will be mined. The network decreases its block subsidy by half after reaching each 210000 block milestone which occurs approximately once every four years. Fidelity states that the halving cycle exists within the system code which enables new blocks to be created approximately every ten minutes through the implementation of difficulty adjustments.

Here is the big idea: each halving removes half of future new supply. So even though 1 million BTC is a small share of the total cap, those last coins come out in smaller and smaller pieces. TradingView said the last full bitcoin could arrive sometime in the 2090s, while tiny fractions will keep being issued until roughly 2140. Investopedia says the next halving is expected in 2028 and will cut the reward from 3.125 BTC to 1.5625 BTC.

This slow tail is one of Bitcoin’s strangest and strongest design choices. A lot of scarce things can become less scarce when people find new deposits or change policy. Bitcoin does not work that way. The issuance path is visible years in advance. That makes supply planning far easier than it is for most assets. Kraken said that with only 1 million coins left spread across more than a century of halvings, supply is now effectively fixed for practical use.

For miners, this matters because future revenue gets tougher with every cycle. In the early years, block subsidy did most of the work. Over time, a smaller subsidy means a bigger push toward lower power cost, better machines, tighter operations, and greater reliance on transaction fees. That does not mean mining stops being useful. It means lazy operations get squeezed out first.

The schedule is easy to see in a halving table. Dates and reward steps below come from Fidelity and Investopedia.

So yes, only 1 million BTC is left. But no, that does not mean Bitcoin is close to the end of mining. It means the remaining supply will be released slowly, in smaller chunks, for a very long time.

Current mining rewards and mining economics in 2026

As of March 17, 2026, the fixed block subsidy is 3.125 BTC. Bitbo listed that figure on its live stats page, and Bitinfocharts showed a recent average reward per block of 3.125 BTC plus about 0.0221 BTC in fees. Bitinfocharts also showed 125 blocks mined in the prior 24 hours, for total block rewards of about 390.63 BTC in subsidy and 2.76 BTC in fees over that stretch.

That sounds like a lot until you zoom out. Mining income is split across a network running at roughly 859 to 872 EH/s depending on which live tracker you look at. Bitbo showed 24-hour hash price near $3.11 per TH per day, while Bitinfocharts showed about $0.0335 per day for 1 TH/s, which points to roughly the same daily economics once you scale it up. In plain terms, every terahash still earns money, but not enough to cover sloppy site design or high power rates.

For Crypto miners, that means profit depends on much more than the block reward itself. Five things drive the business:

• electricity price
• ASIC efficiency in J/TH
• network difficulty
• BTC market price
• pool fees and downtime

When two or three of those move against you at once, payback stretches fast. When they move your way, even one solid machine can do well. This is why many operators care more about power draw per terahash than flashy headline hashrate.

The table below shows the moving parts that matter most. Current reward, difficulty, hashprice, and fee share figures come from Bitbo and Bitinfocharts.

A beginner’s mistake is to look only at machine prices. A better first check is monthly gross revenue, then power bill, then repair risk, then pool fee, then site cooling cost. Mining is not passive. It is an operating business, even when the setup is small.

Top ASIC miners and technical specs

Machine choice matters more now because margins are tighter than they were two or three cycles ago. In 2026, the top SHA-256 machines fall into two broad groups: very efficient hydro units built for stronger sites, and air-cooled units that are easier to deploy but usually a bit less efficient. The right pick depends on your power service, your cooling setup, and whether your location can handle industrial heat and noise.

Bitmain’s S21 XP Hyd is one of the strongest published options in the current market. Bitmain lists it at 473 TH/s, 5676 W, and 12.0 J/T, with 380 to 415V three-phase power and hydro cooling. For operators with the right site build, that is elite efficiency. Bitmain’s air-cooled S21 XP comes in at 270 TH/s, 3645 W, and 13.5 J/TH. The S21 Pro is a step lower at 234 TH/s, 3510 W, and 15 J/TH, still strong for many mid-sized sites and easier to run than a hydro machine.

MicroBT’s current store listing for the WhatsMiner M66S shows a range of 286 to 290 TH/s, 18.5 J/TH, and 5365 W, with 380 to 480V input and a hydro setup. Canaan’s official shop lists the Avalon A15XP at 206 TH/s, 3667 W, and 17.8 J/TH, while the Avalon A15-185T is listed at 185 TH/s, 3420 W, and 18.5 J/TH. Those are not as efficient as the best Bitmain models, but they still belong in the conversation for operators who want more vendor choice or better machine availability.

The table below uses current manufacturer pages or official shop pages. Real site output will vary with temperature, voltage quality, firmware, dust load, and uptime.

When people compare Bitcoin miners in 2026, the smarter question is not “Which box has the biggest TH number?” It is “Which box fits my power, my cooling, and my operating cost?” The best ASIC on paper can still be the wrong buy if your site is not built for it.

Can a normal person still mine Bitcoin in 2026?

Yes, a regular person can still mine Bitcoin in 2026, but the path is narrower than it used to be. Current difficulty is around 145.04 T and live network hash rate is around 859 to 872 EH/s, which makes solo mining with a small setup very hard. Pooling is usually the more practical route because it turns rare jackpot-style wins into steady payouts tied to your share of the pool’s work. Bitbo’s glossary says pools share block rewards according to contributed hash power, and Foundry says its pool uses FPPS payouts.

That does not mean home mining is dead. It means you have to be honest about site limits. Bitmain’s S21 XP and S21 Pro both need 220 to 277V input, pull more than 3.5 kW, and generate strong noise. Bitmain lists 76 dBA for the S21 XP, while Canaan lists up to 75 dB for the Avalon A15 line. Those are not bedroom machines. They are closer to industrial heaters with networking attached.

For smaller Crypto miners, three paths make the most sense:

• one-machine learning setup in a garage, shed, or separate work area
• hosted mining at a site with better power and cooling
• a small rack in a workshop or warehouse with proper exhaust and noise control

Here is a plain comparison. The power, noise, and deployment limits below come from current machine spec pages and pool references.

A normal person can still take part, but the smart move is to start smaller than your ambition. One stable machine teaches far more than ten unstable machines. If your power is high, your room is hot, or your noise tolerance is low, buying a big ASIC first usually ends badly.

Step-by-step guide to start mining Bitcoin

If you want a practical entry path, do not begin with the biggest machine ad you can find. Begin with your site. That one choice shapes almost everything that comes after it. Current Bitmain, MicroBT, and Canaan spec pages make this obvious: voltage range, current draw, operating temperature, and cooling method are not side details. They decide whether a machine can even run safely where you plan to put it.

Step 1: check your power

Look at voltage, breaker capacity, amperage, and uptime. An S21 Pro draws 3510 W. An S21 XP draws 3645 W. The S21 XP Hyd needs 380 to 415V three-phase power. If your site does not match the input spec, stop there and fix that part first.

Step 2: work out your real power cost

Use the number you actually pay after fees and taxes. Mining with a bad power rate can turn a good machine into a bad buy. Bitbo’s live stats show how thin the business can be when hashprice is only around $3.11 per TH per day.

Step 3: pick a machine that fits the site

Air-cooled units are simpler but loud and hot. Hydro units are more efficient but need stronger site planning. If you are brand new, a simpler air-cooled setup is often easier to manage than jumping straight into a hydro loop.

Step 4: choose pool or solo

For most new operators, pool mining is the realistic move. Bitbo says pools make rewards more predictable, and Foundry’s pool docs show how FPPS payouts smooth income by paying for shares plus fee income.

Step 5: set up wallet, pool account, and monitoring

You need a payout wallet, a pool account, worker names, and alerting. Watch hashrate drops, rejected shares, and temperature swings. Downtime is lost revenue that does not come back.

Step 6: build cooling before first power-on

Hot intake air crushes performance and raises failure risk. Manufacturer specs list operating temperature limits, so airflow and exhaust routing have to be part of the plan before the box lands at your door.

Step 7: run a simple payback check

Use three cases: good, average, and rough. Change BTC price, difficulty, and uptime assumptions. If the machine only works in the good case, that is a warning sign. Current hashprice and fee data from Bitbo and Bitinfocharts are useful starting points for this check.

Step 8: scale only after stable operation

Run one unit well for a month or two. Track your actual power draw, room heat, fan performance, pool payouts, and restart behavior. Then add more gear.

Beginner mistakes to avoid:

• buying a machine before checking voltage
• underestimating heat and noise
• ignoring pool fee structure
• using weak internet or unstable power
• treating mining like a set-and-forget side hobby

A careful start is slower, but it saves real money.

Risks, challenges, and mistakes miners should avoid

Mining still works, but it is not forgiving. The first trap is buying hardware that does not fit the site. Plenty of people shop by hashrate and ignore power input, current draw, or cooling type. That is how you end up with a hydro unit at a site with no hydro loop, or a loud air-cooled unit in a room with poor exhaust. Current Bitmain and Canaan pages make it plain that input voltage, temperature range, and noise are not minor details.

The second trap is assuming gross revenue equals profit. It does not. Live March 2026 data shows a network with roughly 145.04 T difficulty and hashprice near $3.11 per TH per day. At those levels, high power cost, downtime, or a poor machine choice can eat most of the margin very quickly.

The third trap is poor operations. Dust, heat, unstable power, bad airflow, and weak monitoring can drag output down without you noticing right away. Many failures do not begin with a dramatic crash. They begin with small drops in hashrate, fan trouble, rejected shares, or thermal stress. Over weeks, those small issues become expensive ones. Manufacturer pages list tight operating ranges for a reason.

Use this table as a reality check. The risks below connect current market conditions with common operating mistakes.

The safest path is boring: stable power, clean airflow, sensible pool choice, strong monitoring, and realistic payback expectations. Mining is still a real business, even if you run only one machine.

What this means for Bitcoin’s future

The 20 million milestone changes the supply story in a very visible way. Bitcoin already had a hard cap before this week, but now the market can see that cap getting close. Kraken said the supply is effectively fixed for practical use. TradingView said the remaining 1 million BTC will take about 114 years to issue. That means fresh supply will keep shrinking while most of the total stock already exists.

That matters for the network too. As block subsidies keep getting smaller, fee income becomes more important. Investopedia says that once all BTC is issued, miners will no longer get block subsidies and will rely on transaction fees. We are not there yet, but the direction is obvious. Every halving pushes the system a little closer to that fee-led world.

It also means the mining business will likely get more selective. Sites with poor power pricing, weak cooling, or older machines will have a harder time staying alive through future halvings. Sites with better engineering, lower power cost, and more efficient gear should keep gaining share. That is not drama. It is just what happens when revenue per block falls on a schedule while competition stays intense. Current March 2026 difficulty and hash rate figures already show how hard that contest is.

A simple big-picture view looks like this:

• less new BTC hits the market over time
• miners care more about fees and uptime
• efficient ASICs matter more with each cycle
• scarcity becomes easier for the market to see

So this milestone is bigger than a round number. It marks a shift in how Bitcoin should be understood: not as a fast-growing new supply asset, but as a capped system where the remaining issuance is small, slow, and highly predictable.