You pay for an order in satoshis. The hashrate you receive mines satoshis back.
Either you end the day with more than you started with or you don't — and
because both sides are denominated in BTC, no price forecast can change the
answer.
Fee assumptions. Both default to zero, the case most
favourable to the marketplace: Braiins does not currently charge the buyer,
and a zero pool fee assumes the bought hashrate is pointed somewhere that
takes nothing. Every real setting moves each figure below down, so a
conclusion drawn at zero is the safe one. For reference, a plain BTC pool
charges 1–2%; unMineable's ADA payout measured ~10% on 2026-08-16.
sat
— on every satoshi committed
Hashvalue
sat per EH/s per day — what the work mines back
Breakeven price
pay more than this and you lose BTC
Top of book
highest resting bid
Delivery in profit
of hashrate being supplied is bought below breakeven, alongside
Each row is a real price quoted by the marketplace right now, measured against
the same hashvalue.
Contracts and Solo packages are sold through the Braiins web interface and are
not exposed by the public API, so they can't be priced live here. Their
arithmetic is identical: divide hashvalue by the quoted price per EH/s per
day. Above sat, the order loses
BTC.
Each dot is a live buy order, placed by what it returns: above the line it
hands back more BTC than it cost, below the line it hands back less. Plotting
return rather than price keeps the good direction up — price would have to run
backwards down the axis to read the same way.
Grey squares are not being supplied. Their return is hypothetical: sellers
will not deliver work for meaningfully less than it mines, so a bid priced under
the line either sits dead or picks up scraps. Anything receiving less than half
the speed it asked for is drawn grey, and the tooltip gives the exact fill —
today's best-looking order is matched at 10% of its own limit. Read these as
prices on offer, not trades available.
The buy side is what people are willing to pay. This is the other half: what
sellers are offering, and how much of it finds a buyer.
Every PH/s that found a buyer sold below hashvalue —
PH/s in total, against
PH/s above the line. Above
hashvalue, PH/s sits unsold:
nobody buys hashrate that mines back less than it costs. Plenty goes unsold
below the line too — buyers are choosy about how far below — but the ceiling
is absolute, and it falls exactly where this dashboard computes it from
network data alone.
Volume-weighted price per 5-minute bar. The marketplace only publishes the most
recent ~8 hours of candles, so this is an intraday view, and the still-forming
final bar is dropped — on partial volume its last print spikes and reads as a
move that didn't happen.
Dashed line is hashvalue ( sat).
Red is where buying
hashpower cost more BTC than it produced;
green is where it cost
less. of
bars traded above the line, below. Weighting by volume matters: the raw last print of each bar straddles
hashvalue roughly half the time, which is noise from a single trade on a thin
book, not a market that is periodically worth buying into.
The fleet is × Antminer S21 Immersion at
TH/s and W
apiece — PH/s drawing
kW at J/TH, on
/kWh power.
Run the fleet
×
—
power costs sat per EH/s/day
Buy the same hashrate
×
—
market asks sat per EH/s/day
Producing hashrate costs % less
than buying it. Same units on both sides, so the two are directly comparable.
Anything left of the dashed line produces BTC for less than it is worth.
Breakeven power price
per kWh, vs contracted —
power could rise
% before mining stops paying
Breakeven BTC price
vs spot — BTC could fall
% before mining stops paying
Breakeven uptime
if a fixed delivery is promised and curtailed hours are bought in at
market, this much self-mining covers the blend
Delivered
PH/s at % uptime
Project net
per day, % share —
site makes
The line is straight, and that is the point. Curtailment removes hashrate and
power in the same proportion, so cost per EH/s/day does not change with
uptime — only the absolute dollars do. Curtailment scales the business
down; it never makes it uncompetitive. The same cancellation applies to the
/ split: because mined BTC and the power bill split on the
same ratio, the project's net is exactly its share of the site's net. The split
decides how the pie is cut, never whether there is one.
Hashvalue is computed from public network data, not from the marketplace being
measured:
Network inputs
| Network hashrate (from difficulty) | EH/s |
| Reward paid per day | BTC |
| Subsidy per block | BTC |
| Average fees per block | BTC |
| Fees as share of reward | % |
Hashrate basis — the denominator is the whole argument
| Difficulty-implied (used) | EH/s |
| Explorer block-arrival estimate | EH/s |
| Estimator divergence | % |
| Hashvalue used | ₿ |
| Hashvalue on the estimate | ₿ |
Difficulty pins hashrate exactly; an estimate reconstructed from observed
block arrivals is a Poisson sample of the same quantity and wanders whole
percent from it for days. The entire order book lives inside ±3% of
hashvalue, so which one you divide by can decide the verdict on the cheap
end of the book. Both are shown so the choice is never silent.
Data as of .