Generating
0 tok/s
last minute
Serving
Generating
0 tok/s
last minute
Capacity
0 tok/s
Requests
0 /min
Nodes
0 online
Memory lent
0 GiB
Served
0 tokens
Tokens per second, one-minute average, last 10 minutes
Machines
| Node | Name | Hardware | Role | Share | Lending | Engine | Online |
|---|
Run hyverk-node with a share of your memory, from 10% to 90%. The hub splits the strongest open model that fits across the machines that are lending. No account: a node is an Ed25519 key created on first run.
Every running node exposes an OpenAI-compatible endpoint on 127.0.0.1:17080. Requests reach the head machine through a signed link; engines never listen on the internet.
The docs are plain Markdown so an agent can set up a node on its own. Start at /llms.txt.