Bitcoin Mining Profitability in August 2026: Hashprice, Break-Even Rates, and What Still Earns

All figures in this analysis are computed from live network data as of August 16, 2026 — a network hashrate of ~860 EH/s, a 3.125 BTC subsidy, and a BTC price near $63,000. They will drift as difficulty and price move; the live Bitcoin calculator always reflects current conditions.

Two months ago, Bitcoin mining looked like an industry in retreat: hashrate had slipped to ~777 EH/s, a ~9% downward difficulty adjustment was landing, and the headlines were all about miners defecting to AI compute. Mid-August tells a more nuanced story. Hashrate has climbed back to roughly 860 EH/s, difficulty followed it upward, and the economics of putting a new machine online are once again a knife-edge question of electricity price. Here is exactly where the numbers stand.

Hashprice: $0.033 per TH/s per Day

The single most useful number in Bitcoin mining is hashprice: how many dollars one terahash per second of computing power earns per day, before electricity. At current conditions it works out to about $0.033/TH/day — the product of the network's 450 BTC daily issuance, a ~$63,000 price, and an 860 EH/s network sharing the pie.

For context, that is meaningfully better than the sub-$0.030 levels reached during the June hashrate trough's aftermath once difficulty re-adjusted, but far below the $0.06+ hashprice that funded the great ASIC build-out of 2024. At $0.033, every machine on the market lives or dies by its efficiency and its power contract.

Break-Even Electricity Rates, Machine by Machine

Multiplying hashprice by a machine's hashrate gives daily revenue; dividing by its daily energy use gives the electricity rate at which it earns exactly zero. Below that rate it profits, above it burns money. Computed from live data:

Three things jump out of that table. First, the efficiency frontier is now the whole game: the 12 J/TH class (S21 XP, XP Hyd) tolerates electricity up to ~$0.11/kWh, while the 23 J/TH S19k Pro — a machine that printed money in 2024 — is now a break-even space heater at $0.06 power. Second, no current machine survives residential European rates: at Germany's ~$0.32/kWh, even the XP Hyd loses over $28 a day. Third, the revenue gap between the M66S and the S21 XP is small enough that acquisition price, not spec-sheet hashrate, should decide the purchase.

The Hashrate Round Trip: June to August

The June story — publicly listed miners re-leasing their power capacity to AI and HPC customers — has not reversed. Core Scientific, Hut 8, and Cipher have all continued expanding GPU hosting through Q3. What changed is the other side of the ledger: the ~9% June difficulty drop mechanically raised per-terahash earnings for everyone who stayed, and that was enough to pull mothballed and relocated capacity back online through July. The result is a network back near 860 EH/s — below January's ~1 ZH/s peak, but well off the June floor.

This is the difficulty mechanism doing exactly what it was designed to do: capacity leaves, margins improve, capacity returns, margins compress again. For small miners, the practical lesson is that windows of elevated profitability after big downward adjustments are short — typically two to four adjustment periods — because the marginal exahash comes back fast.

What This Means If You're Deciding Now

At $0.033 hashprice, the decision tree is short. If your all-in power cost is below $0.06/kWh, current-generation hardware pays for itself in roughly 3–5 years at constant conditions — a real but unspectacular return that depends heavily on BTC price appreciation to beat simply buying the coin. Between $0.06 and $0.10/kWh, only the newest 12 J/TH machines make sense, margins are thin, and a single adverse difficulty epoch can erase a month of profit. Above $0.10/kWh, Bitcoin mining is not an investment; it is a hobby with a utility-bill problem.

Model your own machine, rate, and pool fee with the live Bitcoin mining calculator, compare units side by side in the hardware section, and if your power price is the bottleneck, the electricity-by-country table shows what operators pay elsewhere.