Your payout address is a Quantus address. It starts with qz…. Quantus is post-quantum, so an ordinary Substrate wallet or subkey will not make a valid address. Use one of the two ways below.

Easiest: the official app

1. Install Quantus Wallet (public beta): iOS, Android.
2. Open it and choose Create wallet.
3. Write down the recovery phrase it shows, and keep it offline.
4. Your qz… address is on the main screen. Copy it: that is what you mine to.

Command line: quantus-cli

For a rig with no phone. The tool is called quantus. The ready-made binary is the simplest way and needs no Rust.

1. Ready-made binary. Take the Linux build (…x86_64-unknown-linux-gnu.tar.gz) from the quantus-cli releases page, then:

tar xzf quantus-cli-*-x86_64-unknown-linux-gnu.tar.gz
./quantus wallet create --name pool

2. Or build it with cargo. A fresh Ubuntu has neither Rust nor a compiler, so cargo install quantus-cli on its own will not work. Install both first, then the CLI:

sudo apt update && sudo apt install -y build-essential
curl --proto '=https' --tlsv1.2 -sSf https://sh.rustup.rs | sh
. "$HOME/.cargo/env"
cargo install quantus-cli
quantus wallet create --name pool

create prints your qz… address, which is your payout address. To save the recovery phrase, export it to a file and keep it secret:

quantus wallet export --name pool --format mnemonic --output ./seed.txt

The miner has a benchmark built in. In the folder holding the binary you downloaded on the Start mining tab, run the lines below with the name of the file you downloaded (on Windows, in PowerShell: .\qelvminer-1.0.0.exe):

./qelvminer benchmark --cuda-gpu --cpu-workers 0

It runs for about ten seconds and prints your hashrate. For a steadier number, give it longer. It uses every GPU by default:

./qelvminer benchmark --cuda-gpu --cpu-workers 0 --seconds 30

Qelvminer's tuned engine is CUDA, for NVIDIA cards, and --cuda-gpu turns it on. RTX 30 and 40 series run our own kernels; other NVIDIA cards need NVRTC from the CUDA toolkit (12 or 13). Without the flag, or on other GPUs, the miner uses the Vulkan/DX12 engine, which is much slower. Add --gpu-devices 0,1 to benchmark only some cards (the indices come from ./qelvminer gpu-list).

By default the miner uses every GPU it finds, through the Vulkan/DX12 engine, and every CPU thread but one. The command generated on the Start mining tab already changes both for an NVIDIA rig: --cuda-gpu and --cpu-workers 0. The flags below cap it, split it, or turn one engine off. Every flag also has an environment variable, so it works in a config file or a service unit just as well.

NVIDIA cards: the CUDA engine

--cuda-gpu switches from the Vulkan/DX12 engine to CUDA, which is where Qelvminer is fast. The NVIDIA driver must be installed; without it the miner stops and says so, and you drop the flag.

./qelvminer serve --cuda-gpu   # ...plus your --node-addr / --auth-token / --tls-cert-sha256

GPU only (stop CPU mining)

Frees the processor for the rest of the machine and lets the card do all the work. Set the CPU workers to zero:

./qelvminer serve --cuda-gpu --cpu-workers 0   # ...plus your --node-addr / --auth-token / --tls-cert-sha256

CPU only (no GPU)

Turn the cards off with --gpu-devices none (0 would select card number 0). A CPU hashes a small fraction of what a card does:

./qelvminer serve --gpu-devices none

Set the number of CPU threads

--cpu-workers N caps CPU mining at N threads. Lower it to keep the machine responsive while it mines:

./qelvminer serve --cpu-workers 8

Pick how many GPUs

--gpu-devices 0,2 picks cards by the indices ./qelvminer gpu-list prints; leave it off to use them all. Other GPU knobs: --gpu-throttle-ms N inserts a pause between batches to cut heat and power (0 = flat out), and --allow-integrated lets it use an on-board GPU even when a discrete card is present.

./qelvminer serve --cuda-gpu --gpu-devices 0 --gpu-throttle-ms 5

The same as environment variables

Handy for a .env file or a systemd unit: MINER_CUDA_GPU, MINER_CPU_WORKERS, MINER_GPU_DEVICES, MINER_GPU_THROTTLE_MS, MINER_VERBOSE. Add -v (or MINER_VERBOSE=1) for verbose logs when something looks off.

MINER_CUDA_GPU=true
MINER_CPU_WORKERS=0
MINER_GPU_DEVICES=1
MINER_GPU_THROTTLE_MS=5

Pool or solo

On this pool the mode is decided by the port you connect to, not by a flag: pick Pool or Solo under Mode on the Start mining tab and it sets the right --node-addr: :9834 for PPLNS, :9844 for solo. Solo, on an NVIDIA rig:

./qelvminer serve --node-addr quantus.qelvhash.com:9844 --auth-token <your qz… address>.<worker> --tls-cert-sha256 <pin from the Start mining tab> --cuda-gpu --cpu-workers 0

Barely, and the pool handles it for you. A rig far from the server, in the US or Asia with the server in Europe, loses a sliver of hashrate to the network round trip on every share. The pool measures your connection's round trip and adjusts your share difficulty automatically, giving a distant rig a slightly longer interval between shares so it spends the same tiny fraction on round trips as one next door.

There is nothing to configure, and it works with any miner, ours or the stock one: the pool only changes the difficulty it hands you, so no miner setting or build is involved. Your expected earnings are the same either way.

"timed out" and "reconnecting" in a loop. The pool speaks QUIC over UDP, and some networks block it. Test from the rig with nc -zvu <pool host> <port>; if nothing comes back, open outbound UDP to that port on your firewall or router.

Hashrate drops right after a start, then climbs back. That is the pool's variable difficulty settling on your rig, not a fault. The "Your difficulty" chart on your page shows the ladder moving; give it a few minutes.

The GPU is not used, or the hashrate is far below the benchmarks. Check that the command has --cuda-gpu: without it an NVIDIA card runs the slower Vulkan/DX12 engine. Run ./qelvminer gpu-list to see what the miner detects, then pass --gpu-devices <index> explicitly. On Linux the NVIDIA driver must be loaded (nvidia-smi answers).

"qelvminer only mines on Qelvpool". Qelvminer connects to Qelvpool endpoints only (quantus.qelvhash.com). For another pool, use another miner.

"certificate" or "fingerprint" errors. The miner pins the pool's certificate. Copy the exact --tls-cert-sha256 value from the Start mining tab; a fingerprint from another pool or an old copy will refuse.

A stratum miner. Point it at quantus.qelvhash.com:4444, or at :4443 for stratum over TLS (a self-signed certificate: turn certificate verification off in the miner). The login is your payout address, optionally followed by a dot and the worker name; the password is x. Stratum mines PPLNS; solo is QUIC only, on :9844. With SRBMiner-MULTI, for example:

SRBMiner-MULTI --algorithm quantus --pool quantus.qelvhash.com:4444 --wallet <your qz… address>.<worker>

Your address shows "never sent". The address the rig uses and the one you look up must match character for character. Re-copy it from your wallet; a lookalike character is rejected, never silently corrected.