Solo Mining vs Pool Mining: Which Is More Profitable

Solo and pool mining pay the same expected amount. Your expected reward is your share of network hashrate times what the network pays out, and that formula does not care which one you choose. What changes is when the money arrives: a pool pays small amounts constantly and takes a fee for it, solo pays everything at once or nothing at all. So the honest answer to “which is more profitable” is neither — until you add fees, and then the pool costs you 1-4% while solo costs you the waiting.

The numbers below are from 13 September 2026: a network of about 952 EH/s paying 452.7 BTC a day, with Bitcoin at $77,328.

The expected value is identical, the risk is not

A pool does not pay you more. It pays you the same total, more often, in smaller pieces, minus a fee for the service. Over a long enough stretch — years, not weeks — a solo miner and a pool miner with identical hardware converge on the same gross earnings, because both are paid out of the same network issuance in proportion to the same hashrate.

What pooling changes is the spread of outcomes around that average. Pool mining compresses it: you get something every day. Solo mining leaves it wide open: nothing for years, then an entire block. Two claims often travel together here and only one is true — “pool mining is safer” is correct; “solo mining pays less” is not.

How long a solo block actually takes

Expected time to a block is difficulty × 2³² ÷ your hashrate. At the difficulty of 13 September 2026, that arithmetic produces waits that most people underestimate by an order of magnitude. The last column is the one to read twice: it is the probability of finding anything at all in a month of continuous running.

What it costs to run, either way

Costs are identical whichever you choose — the electricity bill does not know where your shares go. What differs is the payout you are weighing those costs against, so the sensible way to compare is daily revenue against daily cost, for one machine.

One current-generation ASIC at 200 TH/s drawing 3.5 kW, run continuously:

That margin is the whole game, and it is thin enough that the power price decides everything:

• Break-even at 8.8 ¢/kWh if you ignore the hardware and only need to cover power.
• Break-even at 6.6 ¢/kWh if the machine also has to pay for itself over three years.
• Above those numbers you are mining at a loss, and solo mining does not fix it — solo concentrates the same expected revenue into one rare payment instead of smoothing it.

Revenue moves with the Bitcoin price and with difficulty, which adjusts every two weeks. Re-run the first row with today’s figures from our Bitcoin page before making a decision on it.

Fees: the one place the two really diverge

Strip out variance and one real difference remains. A pool takes a cut — commonly around 2% on PPLNS and 4% on the guaranteed schemes — in exchange for aggregating shares and absorbing its own bad luck. A solo miner pays no such fee on the block itself, only whatever the solo service charges when a block is actually found, typically 0-2%.

So over decades, solo keeps a few percent more of the same gross. That is the entire financial advantage, and it is smaller than most people assume when they picture “keeping the whole block”. Which scheme costs what, and who carries which risk, is laid out in our guide to pool payout models.

The threshold where solo starts to make sense

Frame the threshold as a share of network hashrate, not as a number of machines — the network keeps growing, so any fixed count of ASICs goes stale within months while a percentage stays true. Two numbers define the line, and both come straight from the wait-time table above: the share at which a block fits inside a planning horizon, and the funding to sit through it.

Below roughly 0.0002% of network hashrate — about 2 PH/s today, or ten current-generation machines — the expected wait runs past any horizon a small operation can plan around, and the chance of a block inside a year sits near one in ten. Above it, the tail outcome stops being fantasy and becomes a real if unlikely scenario worth planning for.

Two conditions have to hold together before solo is a defensible choice:

1. You can survive the wait. Years of costs with no revenue, funded from somewhere else.
2. You would not be forced to sell the block immediately. If the plan depends on the coin’s price on a specific date, the variance you just bought works against you.

Small miners do find blocks — here is what that proves

Solo services publish the outliers, and they are real. Solo CKPool has found 305 blocks since 2014, several of them for miners running a handful of machines — hashrate far below any sensible threshold.

What those stories prove is exactly the Poisson framing above, not an edge. A miner with a tiny fraction of network hashrate has a small, non-zero chance on any given day; across thousands of such miners over years, some of them win. One documented block does not change the arithmetic — it confirms it.

If you have decided the lottery ticket is worth buying, the fee and the track record of the service you buy it from are the only things left to choose between: our solo pool comparison covers every pool with a real solo mode and what each takes when you win.

What actually changes in the miner

Switching between solo and pool mining is a configuration change, not a hardware one. No reflash, no new firmware, no different machine — you edit four fields in the miner’s web interface and restart it. Two of those fields are harmless to get wrong and two of them can cost you an entire block reward, so they are worth reading slowly.

• Stratum address and port. Solo endpoints are separate addresses from a pool’s shared endpoint, even at the same operator. Copy it exactly; a shared-pool port will not route your work to the solo pot.
• Worker name. Solo services normally use your payout address as the username, so a typo sends a found block’s reward somewhere you do not control, with no recovery. Check it twice.
• Minimum payout. Applies to pool mining; on solo there is nothing to accumulate, the block pays your address directly.
• Maturity. A found block needs 100 confirmations — about 16 hours — before it can be spent. Protocol-enforced, and no service can shorten it.

A straight answer for your setup

Four setups cover almost everyone, and the right answer follows from the wait-time table and the power price rather than from preference:

• One ASIC, or a few at home. Mine in a pool. At 0.1% odds over a month, solo is entertainment, and there is nothing wrong with buying it as entertainment — just do not model it as income.
• A rack, up to a few PH/s. Pool, unless you have outside funding and genuinely do not need the revenue this year.
• Above roughly 2 PH/s with a runway. Solo becomes a real option: the wait is now inside a planning horizon and you keep the few percent the pool would have taken.
• Any size, unstable power price above 9 ¢/kWh. Neither option is profitable at today’s numbers. Fix the power price first; the payout model cannot save an operation that loses money per kilowatt-hour.

FAQ

Sources

• mempool.space — network hashrate — hashrate and difficulty used for every wait-time calculation, 13 September 2026
• mempool.space — reward stats — daily issuance including transaction fees
• mempool.space — prices — Bitcoin price used in the cost table
• Solo CKPool — all-time solo block count and fee
• mempool.space — Solo CK attribution — independent confirmation of blocks found