Antminer L7 vs L9 vs L11
| Machine | Hashrate | Wall | J/MH | kWh/day | Gross $/day | Net at your rate | Break-even $/kWh |
|---|---|---|---|---|---|---|---|
| Antminer L7 | 9.5 GH/s | 3,425 W | 0.36 | 82.2 | $6.06 | −$2.16 | $0.074 |
| Antminer L9 | 16 GH/s | 3,360 W | 0.21 | 80.6 | $10.20 | +$2.14 | $0.127 |
| Antminer L11 | 20 GH/s | 3,680 W | 0.184 | 88.3 | $12.75 | +$3.92 | $0.144 |
| Antminer L11 Pro | 21 GH/s | ~3,612 W* | 0.172 | 86.7 | $13.39 | +$4.72 | $0.154 |
Gross = live combined LTC+DOGE hashprice × hashrate, before pool fee. *L11 Pro wall watts are nominal (derived from 21 GH × 0.172 J/MH); Bitmain tolerances are ±3% hashrate, ±5% power on all machines. Static $ figures dated 2026-09-25; live values replace them at load.
The only three numbers that differ
Since revenue per GH is identical everywhere (that's the definition of hashprice — both legs of it, LTC plus merged DOGE), the whole comparison reduces to: how many GH you get, what each costs to run (J/MH), and what the machine costs you. Two of those are on the spec sheet; the third we don't quote — street prices rot faster than pages.
Machine by machine
L7 — the exposed veteran. 9.5 GH/s at 3,425 W is 0.36 J/MH, which means paying about 71% more for electricity per hash than an L9 and roughly twice an L11 Pro's rate. It ships in factory speed bins (locked at manufacture) from 8.8 to 9.5 GH, and the used market is dominated by the 9.3 GH bin — verify the bin before you pay for the top one. On cheap power it still pencils; the live L7 page computes exactly where.
L9 — the mid path and halving survivor. 16 GH/s at 3,360 W (0.21 J/MH), with a 17 GH/3,570 W variant. The cleanest receipt in scrypt: the 17 GH L9 draws about 4% more watts than an L7 while hashing 79% more — same plug, same circuit, nearly double the hash. No longer the efficiency frontier, but efficient enough to cross August 2027 at most non-residential rates. Live page.
L11 and L11 Pro — the current frontier. 20 GH/s at 0.184 and 21 GH/s at 0.172 J/MH — the Pro buys ~5% more hash and ~7% better efficiency over the base. This class has the most post-halving headroom, and the used market prices that in. L11 · L11 Pro.
Which one for your situation
Cheap or industrial power: electricity is a small share of revenue, so older iron can clear its penalty — the L7's 71% energy-per-hash overhead on a small cost base can still work. Run your actual rate in the table above; don't take a blog's word for it, including ours.
Garage on residential 240V: power dominates the math — L9 class is the floor, L11 class if the purchase price works.
Apartment: none of these — they're 200–240 V datacenter machines in the ~75 dB(A) class (manufacturer claims); the honest options are the home-class scrypt machines: Mini-DOGE III, DG Home 1, Fluminer L1 Pro.
Buying used: a cheap used L7 is cheap because the market has already priced 0.36 J/MH against August 2027 — you're buying a shorter runway, not a discount on the same asset. Verify hashboards, fans and the factory bin before paying.
The August-2027 cutoff
At block 3,360,000 — 176,027 blocks (~306 days) from the last fetch — the LTC subsidy drops from 6.25 to 3.125. Not “revenue halves”: the DOGE side's fixed 10,000 per block keeps paying, fees continue, so revenue per GH falls by less than half. What the halving actually does is move every machine's break-even line down — and machines shut off in efficiency order, worst J/MH first: L7 class before L9, L9 before L11. When each crosses depends on your rate and on market prices, which is exactly what we don't predict; the countdown runs from our own node's height, and the table computes your case today.
Circuit and room reality
| Machine | Input | Amps @ 240 V | Heat output | Noise (mfr claim) |
|---|---|---|---|---|
| L7 | 200–240 V | ~14.3 A | ~11,700 BTU/hr | ~75 dB class |
| L9 (16 GH) | 200–240 V | ~14.0 A | ~11,500 BTU/hr | ~75 dB class |
| L11 | 200–240 V | ~15.3 A | ~12,600 BTU/hr | ~75 dB class |
| L11 Pro | 200–240 V | ~15.1 A (nominal) | ~12,300 BTU/hr | ~75 dB class |
None of these runs on a standard 120 V/15 A household circuit, and at ~15 A they want a dedicated 20 A branch at 240 V under the 80% continuous-load rule. The heat is a space heater's; the noise is a shop vac's that never stops. This section disqualifies more buyers than the economics do — by design, before money is spent.
FAQ
- Is the Antminer L7 still worth running in 2026?
- Only on cheap power — its 0.36 J/MH pays ~71% more per hash for electricity than an L9, and August 2027 cuts the LTC side of its revenue in half. Your rate in the table above gives the verdict, not our opinion.
- What's the actual difference between the L9 and the L11?
- Hashrate and efficiency, nothing exotic: 16–17 GH at ~0.21 J/MH vs 20–21 GH at 0.184–0.172. Revenue per GH is identical; whether the L11 beats a discounted L9 depends on asking prices, which we don't quote.
- Which Antminer should I buy for Dogecoin?
- Any of these — all three merged-mine LTC and DOGE simultaneously at full hashrate. There is no DOGE-only Antminer to pick.
- What happens at the 2027 halving?
- LTC subsidy 6.25 → 3.125; DOGE and fees keep paying, so revenue per GH falls by less than half — and machines shut off worst-J/MH-first. When depends on your rate; the table computes your case.