A pool for the AI compute revolution.
modelOS turns every GPU forward pass into verifiable proof of work. Mine MDL and PRL simultaneously — one miner, two rewards, zero extra config.
The NetworkLive state of the pool
Pool Hashrate
—
over last 4 min
Network Hashrate
17.39 EH/s
current difficulty
Pool Share
0.00%
of network hashrate
Active Miners
0
in PPLNS window
Blocks Found
8,694
all-time
Total Paid MDL
250.79K MDL
distributed to miners
Block height: 28,380Pool fee: 2%PPLNS window: 50000 sharesAuxPoW: MDL + PRL
On-ChainPool wallet activity
Blocks Found
8,694
Total MDL Rewarded
258.41K MDL
Total PRL Rewarded
0 PRL
via AuxPoW distribution
Current Block
28,380
Begin Mining
One command. Then your GPUs are mining MDL.
Run it on Docker, a rented GPU (RunPod / Vast.ai), or HiveOS — paste your address and go. Rewards distributed automatically after maturity. Add a Pearl address to also earn PRL via AuxPoW.
Configuration
Start mining
Required — block rewards paid here
Add your Pearl address to also earn PRL via AuxPoW — the pool handles the rest
Own NVIDIA GPU? One command — needs Docker + the NVIDIA Container Toolkit.
docker run -d --restart unless-stopped --gpus all --shm-size 8g \ -e POOL_HOST=stratum.modeloslab.xyz \ -e POOL_PORT=5566 \ -e POOL_TLS=false \ -e MDL_ADDRESS=mdl1p… \ -e POOL_WORKER=rig-01 \ -e HF_TOKEN=hf_… \ -e MODELOS_MODEL=8b \ -e MODELOS_MODEL_VERSION=4 \ -e MODELOS_TENSOR_PARALLEL_SIZE=1 \ -e MODELOS_GPU_INDICES=0 \ -e VLLM_PORT=8000 \ -e MINER_NO_GATEWAY=true \ -e MINER_NO_VLLM_PLUGIN=true \ -e MODELOS_GPU_MEMORY_UTILIZATION=0.60 \ -e MODELOS_MAX_MODEL_LEN=8192 \ -e MODELOS_COORDINATOR_URL=https://pool.modeloslab.xyz \ -e MODELOS_WORKER_PRIVKEY=<worker-private-key> \ -e MODELOS_DIRECT_MINER_V2=1 \ -v modelos-models:/model-cache \ modeloslab/vllm-miner:latest
stratum.modeloslab.xyz:5566 (PPLNS) · stratum.modeloslab.xyz:5567 (SOLO) · 2% fee
VRAM: inference + mining needs room for the model and the ~6 GB the proof reserves (fixed — it can't be reduced). A 32 GB RTX 5090 is the lowest GPU that fits a model alongside the miner — it runs the 8B tier. 24 GB (4090) and 8 GB (3070) cards must use Miner only.
| 8B | ~16–18 GB | RTX 5090 32 GB (lowest), A100, H100 |
| 14B | ~28 GB | A100 40/80 GB, H100, 2×24 GB (TP=2) |
| 32B | ~64 GB | A100 80 GB, H100, multi-GPU (TP) |
| 70B | ~140 GB | 2×80 GB A100/H100 (TP=2) |
Tune
MODELOS_GPU_MEMORY_UTILIZATION: 0.60 is conservative — it leaves the miner ~6 GB on a 32 GB card. Raise iton larger cards (A100/H100) or multi-GPU rigs to fit bigger models; lower it if vLLM OOMs. For multi-GPU, set MODELOS_TENSOR_PARALLEL_SIZE to the number of cards the model spans.Inference FeesTwo income streams per GPU.
Block Rewards
Win the PoW race
Every NoisyGEMM forward pass generates verifiable proof of work. When your hash meets the target, you earn the MDL block reward. AuxPoW shares also earn PRL from the Pearl upstream pool.
Rewards are split proportionally via PPLNS and distributed after 100-block maturity on both chains.
Rewards are split proportionally via PPLNS and distributed after 100-block maturity on both chains.
Inference Fees
Win the inference race
Run vllm-miner and compete for paid AI inference jobs. When an
inference_tx confirms on-chain, all vllm-miners race to execute the prompt and submit a cryptographic proof. First confirmed proof earns the locked fee — no pool cut.| DeepSeek R1 0528 (8B) | 0.005 MDL base |
| Qwen3 14B | 0.02 MDL base |
| Qwen3 32B | 0.05 MDL base |
| DeepSeek R1 70B | 0.10 MDL base |
Lookup
Check your earnings.
Enter your MDL wallet address to see your hashrate, workers, and payment history in real time.