Avalon Nano 3S review 2026: is the 6 TH/s home miner worth it?
By Lukas Henning · 26. August 2026 · 12 min read
The Avalon Nano 3S by Canaan is a desktop miner that delivers roughly 6 TH/s at about 140 watts and stays quiet enough for a living room. The power supply is built in, setup runs through a web interface without a cloud account, and after that it needs nothing but power and a network connection. For a box the size of a thick book, that is a lot of hashrate.
This review puts the numbers into context: how much electricity it actually draws, how loud it is in each mode, what the block odds look like when solo mining, and where it wins or loses against open alternatives such as the Bitaxe Gamma, Nerdaxe Gaia and NerdQaxe++.
The Nano 3S packs a lot of hashrate into a box you plug in and then leave alone. The catch is efficiency. Per terahash it needs roughly twice the energy of a current open-source miner such as the Bitaxe Gamma or the Nerdaxe Gaia, and electricity is the bill that comes back every month.
Avalon Nano 3S at a glance
| Hashrate | ~6 TH/s (high) · 4.8 TH/s (medium) · 3.1 TH/s (low) |
| Power draw | ~140 W · 104 W · 66 W depending on mode |
| Efficiency | ~23 J/TH in high mode |
| Noise | ~29 to 36 dB depending on mode (some tests measure up to 40 dB) |
| Firmware | Canaan firmware, not open source |
| Network | 2.4 GHz WiFi and LAN (RJ45) |
| Solo mining | yes, any pool can be entered (Stratum V1) |
| Power supply | built in (100 to 240 V) |
| Best for | beginners who want hashrate without tinkering |
Our rating: 8.0 / 10. Points for the hashrate in a small case, the noise level and a setup that takes ten minutes even without prior knowledge. Points off for efficiency (23 J/TH is well behind the roughly 10 J/TH of a Nerdaxe Gaia) and for the closed firmware.
Avalon Nano 3S technical specifications
| Specification | Value |
|---|---|
| Manufacturer | Canaan (Avalon) |
| Algorithm | SHA-256 |
| Coins | Bitcoin (BTC), technically BCH/BSV as well |
| ASIC | 12× Canaan A3197 at 4 nm, the same chip family as the larger Avalon 1566 |
| Hashrate | 6 TH/s in high mode (manufacturer figure) |
| Power draw | 140 W (high), 104 W (medium), 66 W (low) |
| Efficiency | ~23.3 J/TH (high), ~21.7 (medium), ~21.3 (low) |
| Work modes | three, switchable while running |
| Noise | ~29 dB (low) to ~36 dB (high) per manufacturer |
| Dimensions | approx. 20.5 × 11.5 × 5.8 cm |
| Weight | approx. 0.7 to 0.86 kg (sources differ) |
| Network | 2.4 GHz WiFi and LAN (RJ45) |
| Pool | freely configurable, Stratum V1 |
| Firmware | Canaan firmware (closed source) |
What the Avalon Nano 3S is like day to day
Setup. Unbox, plug in, open the web interface, enter WiFi and pool. The flow is built so that nobody has to flash firmware. Anyone who has set up a Bitaxe will notice the difference straight away: fewer options, and fewer ways to get something wrong.
Heat. All 140 watts end up as heat in the room, about as much as a laptop under full load. Pleasant in winter, noticeable in a small room in high summer. Canaan markets the Nano line as a small heater as well. It should still stand in the open: in a closed cabinet it heats up, throttles and the fan gets louder.
Continuous operation. The device is built for 24/7 operation. Some test reports mention isolated cases of overheated USB connectors. It does not look like a widespread problem, but it is a good reason to stick with the supplied power supply and cable.
How loud is the Avalon Nano 3S?
Quiet enough for a living room, too loud for a bedroom if you are a light sleeper. The published figures diverge, and there is a reason: almost no source states the measuring distance.
| Source | Figure |
|---|---|
| Canaan / retail listings | 29 dB (low) to 36 dB (high) |
| Independent test (blockdyor) | 33 to 36 dB, own measurement |
| Hashrate Index | ~33 dB (low) to ~40 dB (high) |
The spread comes almost entirely from mode, measuring distance and the room’s own noise floor.
For reference: 30 dB is a very quiet room, 40 dB a soft conversation next door.
- Bedroom: not in high mode. In low mode, tolerable for many people.
- Office or living room: unobtrusive, as long as it is not right next to your head.
- Basement or utility room: completely unproblematic.
A Bitaxe Gamma is audibly quieter; at 15 to 18 watts its fan barely has to turn. A NerdQaxe++ at 80 watts sits in between. If quiet matters most, run the Nano 3S in low or medium mode: you lose hashrate, and the efficiency even gets slightly better.
Power consumption and electricity cost
The formula is simple: watts ÷ 1000 × 24 × 30 × price per kWh gives the monthly cost. High mode draws 140 W, which makes 100.8 kWh per month and 1,226 kWh per year.
| Electricity price | per month | per year |
|---|---|---|
| €0.25/kWh | €25.20 | €306.60 |
| €0.30/kWh | €30.24 | €367.92 |
| €0.35/kWh | €35.28 | €429.24 |
| €0.40/kWh | €40.32 | €490.56 |
The work modes change that considerably, calculated at €0.30/kWh:
| Mode | Hashrate | Power | per month | per year |
|---|---|---|---|---|
| Low | 3.1 TH/s | 66 W | €14.26 | €173.45 |
| Medium | 4.8 TH/s | 104 W | €22.46 | €273.31 |
| High | 6.0 TH/s | 140 W | €30.24 | €367.92 |
Low mode at 21.3 J/TH is actually more efficient than high mode at 23.3 J/TH, so you pay disproportionately for the last 1.2 TH/s. If the electricity bill matters to you, run medium: 80 percent of the hashrate for 74 percent of the power. Run your own numbers in the mining calculator.
Efficiency: is 6 TH/s at 140 watts good?
J/TH (joules per terahash) tells you how much energy a miner needs for the same work. Lower is better. Over the life of a miner this number decides more about your cost than the purchase price does.
| Device | Hashrate | Power | J/TH | Electricity per year at €0.30/kWh |
|---|---|---|---|---|
| Avalon Nano 3S | 6.0 TH/s | 140 W | 23.3 | €367.92 |
| Bitaxe Gamma 601 | 1.3 TH/s | ~16.5 W | 12.7 | ~€43 |
| Nerdaxe Gaia | 2.6 TH/s | 25 W | 9.6 | €65.70 |
| NerdQaxe++ | 4.8 TH/s | 80 W | 16.7 | €210.24 |
The Nano 3S brings the most hashrate of the four devices, and it is also the least efficient. Compared with the Nerdaxe Gaia it needs about 2.4 times the energy per terahash.
So it makes sense to look at both together. For roughly equal hashrate you would need two Gaia: 5.2 TH/s at 50 watts. They use about €132 of electricity a year instead of €368, which saves around €236 a year or €20 a month. Take the difference in purchase price between one Nano 3S and two Gaia from the product pages, divide it by €20, and you have the number of months after which the more efficient option has paid for itself.
More TH/s therefore does not automatically make a miner the better choice. For an hour a day efficiency barely matters; around the clock it decides almost everything. All devices side by side are in the miner comparison.
Solo mining with the Avalon Nano 3S
In solo mining you do not work for a pool that splits rewards by share. You try to find a complete block yourself. If you do, the full reward is yours: currently 3.125 BTC plus fees. If you do not, you get nothing.
Every hash is its own lottery ticket, and the probability rises linearly with hashrate: 6 TH/s is about 4.6 times as many attempts as the 1.3 TH/s of a Bitaxe Gamma.
With network data from 26 August 2026 (difficulty 125.807 T, network hashrate 893.6 EH/s), a block takes 5.4 × 10²³ hashes on average. For 6 TH/s that gives:
| Period | Probability |
|---|---|
| per day | 1 in 1,042,316 |
| per month | 1 in 34,743 |
| per year | 1 in 2,853 |
Our own calculation based on the difficulty from mempool.space, as of 26 August 2026.
For comparison, the same calculation gives roughly 13,170 years for a Bitaxe Gamma, 6,585 for a Nerdaxe Gaia, 3,567 for a NerdQaxe++ and 1,223 years for a NerdOctaxe Hydro at 14 TH/s.
At €368 a year and annual odds of 1 in 2,853, the electricity is the price of a ticket in a very large lottery, and it belongs in the hobby budget.
Can the Avalon Nano 3S find a whole Bitcoin block?
Yes, technically it can. There is no lower limit below which a miner would be "too small" for a block.
A valid block exists when any hash falls below the current target. Whether that hash comes from a farm with 100,000 machines or from the Nano 3S on your desk makes no difference to the network: the block is accepted under the same rules and you receive the full reward.
Size only changes the number of attempts. At 6 TH/s the chance is about 1 in 1,042,316 per day. Small, but not zero, and that is the appeal for many home miners. For this to work you need to mine on a solo pool that pays the block reward to your own address. pool.opensourceminers.de is one such solo pool and works without registration.
Avalon Nano 3S vs Bitaxe Gamma
| Avalon Nano 3S | Bitaxe Gamma 601 | |
|---|---|---|
| Hashrate | 6 TH/s | 1.3 TH/s |
| Power | 140 W | ~15 to 18 W |
| Efficiency | 23.3 J/TH | 12.7 J/TH |
| Electricity per year (€0.30/kWh) | €367.92 | ~€43 |
| Firmware | Canaan, closed source | AxeOS, open source |
| Block odds / year | 1 in 2,853 | 1 in 13,171 |
| Best for | plug and play | learning, tinkering, efficiency |
The Bitaxe Gamma is the better pick if you want to understand how mining works, if the device sits next to your bed, or if low running costs matter more to you than absolute hashrate. The Avalon Nano 3S suits you if you want 4.6 times the block odds without touching firmware.
Avalon Nano 3S vs Nerdaxe Gaia
The Nerdaxe Gaia is the more interesting opponent: its much newer BM1373 ASIC gives it the best efficiency of any miner in our range, around 10 J/TH.
| Avalon Nano 3S | Nerdaxe Gaia | |
|---|---|---|
| Hashrate | 6 TH/s | 2.6 TH/s |
| Power | 140 W | 25 W |
| Efficiency | 23.3 J/TH | 9.6 J/TH |
| Electricity per year (€0.30/kWh) | €367.92 | €65.70 |
| Firmware | Canaan, closed source | NerdOS (AxeOS fork), open source |
Over three years of continuous operation at €0.30/kWh, the Nano 3S uses about €1,104 of electricity for 6 TH/s. Two Gaia deliver 5.2 TH/s and use about €394 in the same time. The gap of roughly €710 is far larger than the difference in purchase price, which you can check on the two product pages.
This is where the Nano 3S loses most clearly. The closed firmware is a matter of taste; the electricity bill arrives every month.
Avalon Nano 3S vs NerdQaxe++
| Avalon Nano 3S | NerdQaxe++ | |
|---|---|---|
| Hashrate | 6 TH/s | 4.8 TH/s |
| Power | 140 W | 80 W |
| Efficiency | 23.3 J/TH | 16.7 J/TH |
| Electricity per year (€0.30/kWh) | €367.92 | €210.24 |
| Firmware | Canaan, closed source | NerdOS (AxeOS fork), open source |
| Block odds / year | 1 in 2,853 | 1 in 3,567 |
The Nano 3S is 25 percent faster. The NerdQaxe++ saves about €158 of electricity a year, roughly €13 a month. If the Avalon is the cheaper device on the day you buy, the NerdQaxe++ earns that difference back at about €13 a month in continuous operation. After that, the Avalon is the more expensive machine to own.
Medium mode changes the picture: there the Nano 3S delivers the same 4.8 TH/s as the NerdQaxe++ but needs 104 instead of 80 watts. The electricity gap shrinks to about €63 a year, so switching modes recovers a good part of the efficiency difference.
Open source or closed firmware: does it matter?
Yes, but probably not where you would first expect.
What open firmware gives you: the source code is readable, so you can see what the device does and where it sends data. You can set frequency and voltage yourself and shift the operating point between efficiency and performance. Pools are freely configurable, updates come from the community, and schematics are documented if something fails. If the manufacturer disappears, the device keeps running.
What the Avalon gives you: a finished product. The setup walks you through the few settings, the modes are tuned to each other, the power supply fits. You can get less wrong because there is less to set.
Both are reasonable answers to different questions. What you give up with the Avalon is the fine tuning that often gets another 10 to 20 percent of efficiency out of open devices, and the option to keep the machine running on new firmware five years from now.
Who is the Avalon Nano 3S right for?
Buy the Avalon Nano 3S if …
- you want a lot of hashrate in one plug-and-play box and efficiency is secondary
- you have no appetite for firmware, tuning and troubleshooting
- the device will only run in winter anyway and the waste heat is welcome
- you have cheap or self-generated electricity; with surplus solar the whole calculation changes
- you are looking for a gift that works without a manual
Take an open miner instead if …
- the device runs 24/7 on a normal household tariff, where efficiency decides
- you mine in the bedroom and every decibel counts
- you want to understand and adjust what is happening
- you want to keep the option of modifying or repairing the device later
- you want to start on a small budget
Pros and cons
For
- 6 TH/s in a case the size of a thick book
- more TH/s per euro than the three open miners in this comparison (current prices on the product pages)
- three power modes, switchable while running; low is even more efficient than high
- power supply built in, no accessories to buy
- setup in under ten minutes, no prior knowledge needed
- freely selectable pool, solo mining possible
- quiet enough for living spaces
Against
- 23.3 J/TH, roughly 2.4 times the energy per terahash of a Nerdaxe Gaia
- about €368 of electricity per year at €0.30/kWh, more than the purchase price
- closed firmware: frequency and voltage cannot be tuned, and the code cannot be inspected
- partly contradictory manufacturer figures (29 J/T versus 23.3 J/TH)
- isolated reports of overheated USB connectors
Plug-and-play hashrate with a high electricity bill
The Avalon Nano 3S is the right choice when as much hashrate as possible in an uncomplicated device matters more to you than efficiency and open software. You get a lot of block odds for the money, a setup anyone can manage, and a machine that does not stand out in a living room.
The price for that shows up on the electricity bill. At 23.3 J/TH it is among the least efficient home miners of its generation, and in continuous operation it costs you more in electricity every year than the device itself. If you mine around the clock on a normal tariff, a Nerdaxe Gaia or a NerdQaxe++ works out considerably cheaper over three years once electricity is included.
Our advice: if the device only runs part of the time, in winter or on your own solar power, the Nano 3S is a good deal. If it runs around the clock on a grid tariff, buy efficiency instead of hashrate. And if you are unsure, run it in medium mode: 4.8 TH/s at 104 watts is the most sensible point on its curve.
Frequently asked questions
How many TH/s does the Avalon Nano 3S have?
How much power does the Avalon Nano 3S use?
How loud is the Avalon Nano 3S?
Can the Avalon Nano 3S do solo mining?
Can the Avalon Nano 3S find a Bitcoin block?
Is the Avalon Nano 3S worth it?
Avalon Nano 3S or Bitaxe?
Avalon Nano 3S or NerdQaxe++?
Is the Avalon Nano 3S open source?
Which chip is inside the Avalon Nano 3S?
Written by
Lukas Henning · Mining editor and hardware expert
Lukas has been working with Bitcoin mining for years and runs several open-source miners such as Bitaxe and NerdQaxe at home. For Open Source Miners he tests hardware and turns his setups into guides you can follow step by step on your own device.
All guides by Lukas Henning →