Serve AI. Mine $PEARL.
Same compute, paid twice. Peryn maximizes the yield of every GPU.
- Start mining
- 1
- command to start mining
- GPUs supported
- 64
- GPU models supported
- Stored prompts
- 0
- prompts or answers stored
- GPU hand-back
- <1
- second to hand a GPU back
- Pool hashrate
- 2.20 EH/s
- Miners online
- 4,208
- Blocks found
- 22
- Live
- updated 20:12 UTC
Fees: 1% on PPLNS, 1% on PPS+ (the default).
For miners: $PEARL to your own address.
Mine on NVIDIA GPUs from Turing to Blackwell and earn $PEARL to your own address. GPUs that serve AI earn USD at the same time.
- $PEARL to your address
- Paid every 2 hours to your $PEARL address once you are owed 1 $PEARL or more.
- Simple setup
- One command on Slurm, bare metal or any mining OS.
- Every share shown
- See every share and block on your dashboard.
- Pool terms
- Terms set in our pool terms.
Your GPUs earn USD and $PEARL at once.
Serve AI for USD. With our datacenter runtime your GPUs also mine $PEARL: inside serving, between requests and when idle. Your scheduler keeps every GPU it asks for.
-
Hopper and Blackwell
Datacenter and workstation cards. See the list
-
Your scheduler wins
Peryn hands the GPU back in under a second once your scheduler asks.
-
Your customers’ data
Never leaves the machine.
-
Every setup
One command on Slurm, bare metal or any mining OS.
One GPU. Two incomes.
Every AI job on Peryn also mines $PEARL, inside the work and in the gaps. Customers pay less per token; GPU owners earn more.
-
You send a request
Through our API or the chat, to one of four open models.
-
A GPU answers it
On a node in our network, checked with hidden test prompts.
-
A second income
While it answers, between requests and when idle, it mines $PEARL.
Use AI. Private by default.
Four top open models, one OpenAI-compatible API. We never store prompts or answers, and we never train on them.
Talk to us.
Private tier, monthly invoicing, volume pricing or a model you need. We reply within one business day.