Tool guide
How long until your machines pay for themselves
Compare when a fleet could recover its initial cost under three difficulty-growth assumptions, with the model’s limits shown.
Doefin Research ·
The model starts with your initial outlay, then adds mining revenue and subtracts power costs each month. It shows the first month that cumulative modelled cash flow reaches zero, if that happens within your chosen period. Compare three assumed difficulty-growth rates to see how the result changes.
Use this if you own mining machines and want to compare when their initial cost could be recovered under stated assumptions. Hosting operators can model fleets they own. Fee-only hosting, bought hashrate, non-mining income and a full investment valuation need another model.
How to use it
- Enter Fleet hashrate in TH/s and Efficiency in J/TH for the same fleet and operating mode.
- Enter All-in power cost per kWh and What you paid up front. Check which costs they cover; the model adds no omitted costs elsewhere.
- Choose a constant Bitcoin price and enter a monthly Difficulty growth guess. The default 1.5% is an illustrative input, not a site forecast.
- Check Starting difficulty and Model horizon. The URL also accepts diffT and months within the input limits. Defaults are 127.48T and 48 months.
- Compare the central result with lower and higher growth. Lower is your rate minus two percentage points, with a minimum of zero; higher is your rate plus three points.
- Read the final cash balance as well as the payback month. Payback marks the first month the model recovers the initial cost; the balance shows where it ends.
- Copy the URL to save the inputs. Anyone with the link can read your assumptions.
What each setting means
Every number the calculator asks for, what it is, and where to find yours.
Fleet hashrate
Combined modelled fleet hashrate in TH/s.
Where to find it: Use fleet specifications or an operating average for the same machines. Record whether that hashrate already allows for downtime.
Input limits and sample: The input accepts 100–20,000 TH/s. The illustrative default is 1,000 TH/s.
Efficiency
Joules per terahash. Hashrate × efficiency determines modelled wall power.
Where to find it: Use a measurement or specification relevant to the same operating mode as your hashrate.
Input limits and sample: The input accepts 12–40 J/TH. The illustrative default is 15 J/TH.
All-in power cost
The chosen cost per modelled kWh. No separate operating-cost input is added.
Where to find it: Use your billing terms. If you convert another cost to $/kWh, record the approximation and do not count it twice.
Input limits and sample: The input accepts $0.02–$0.15/kWh. The illustrative default is $0.045/kWh, not a market quote.
What you paid up front
The amount paid up front that the fleet needs to recover. Changing hashrate does not change this input.
Where to find it: Use invoices or a stated scenario amount. The model does not validate equipment prices or purchase history.
Input limits and sample: The input accepts $5,000–$2,000,000. The illustrative default is $20,000.
Bitcoin price
A constant USD price used to value all modelled mined BTC.
Where to find it: Choose and record a scenario price. It does not predict a future price path.
Input limits and sample: The input accepts $20,000–$200,000 per BTC. The illustrative default is $90,000 per BTC.
Difficulty growth guess
The monthly difficulty increase, applied repeatedly. The first increase happens before the model calculates production for month one.
Where to find it: Choose a sensitivity assumption. Historical observations are not a guarantee of a future growth rate.
Input limits and sample: The central input accepts 0–8% per month. Its illustrative default is 1.5%. The comparison subtracts two percentage points, with a minimum of zero, or adds three.
Starting difficulty
The starting network difficulty in trillions. It sets the starting point for calculating expected production.
Where to find it: Check a block-data source you trust and enter its value in Starting difficulty.
Input limits and sample: The input accepts 1T–1,000T. The illustrative default is 127.48T, not current network difficulty.
Model horizon
The number of months to calculate before the model ends. Each month has 30.4 modelled days.
Where to find it: Choose the period over which you want to test recovery of the initial outlay. A result ending below zero does not establish what happens after that period.
Input limits and sample: The input accepts 2–120 months. The illustrative default is 48 months.
Reading the answer
Payback
The first month when mining revenue, after power costs, has recovered the initial outlay in the central scenario.
So: This estimate depends on your inputs and leaves out costs outside them. It does not guarantee that you will recover the outlay.
The three growth rows
Lower, central and higher monthly difficulty-growth assumptions for the same fleet and costs.
So: These scenarios show how a change in assumed difficulty growth changes payback. They are not a confidence interval, forecast or investment ranking.
Cash after the modelled horizon
All modelled mining revenue, minus power costs and the initial outlay. Mined BTC is valued at your chosen constant price.
So: A negative value is a deficit at that date. It cannot prove payback will never occur.
Power cost per day
Hashrate × efficiency ÷ 1,000 × 24 hours × the power-rate input. For the sample this is $16.20/day.
So: The model holds it constant and does not automatically add other operating costs.
Reproduce the example under three growth assumptions
Use 1,000 TH/s, 15 J/TH, $0.045/kWh, $20,000 outlay, $90,000/BTC, 127.48T, 1.5% central growth and 48 months. Inputs are illustrative.
- The fleet draws 15 kW, consumes 360 kWh per modelled day and costs $16.20/day in power.
- At 0% monthly difficulty growth, the first non-negative cumulative cash month is 24. The final 48-month balance is approximately +$21,123.
- At 1.5% monthly growth, the first non-negative cumulative cash month is 38. The final balance is approximately +$2,291.
- At 4.5% monthly growth, there is no payback within 48 months. The final balance is approximately −$17,282.
The same fleet reaches different results under each assumed difficulty path. These estimates leave out parts of an investment valuation and do not calculate a contract position size.
Mistakes people make
- Interpreting no payback within 48 months as a claim about every later year.
- Treating sample equipment cost, BTC price or growth as a current market benchmark or recommendation.
- Changing hashrate while unintentionally retaining an unrelated capital outlay or power assumption.
- Ignoring costs, downtime or additional income that the simplified model does not represent.
What it does not tell you
- The subsidy is fixed at 3.125 BTC throughout the horizon, including across a future halving. The model cannot represent an actual horizon crossing a subsidy change.
- Transaction fees, pool fees, downtime, repairs, tax, financing, resale value and non-mining income are not separate inputs. Costs folded into the power rate are approximations.
- BTC price, hashrate, efficiency and power rate stay constant. Difficulty follows a compounded constant monthly assumption, not an observed or forecast path.
- Each month has 30.4 modelled days. Starting difficulty grows once before month-one revenue is calculated.
- The lower-growth path is floored at zero. Falling difficulty cannot be modelled. Three paths do not bound every possible outcome.
- A binary difficulty contract observes one specified block and has its own position cost, fees and payout in its collateral. You can lose the full amount paid for a position. This model contains no such position and does not assess hedge economics.
