$hanzo fabric up
the level of the work
specialized compute like proofs
Consumer cardsdatacenter acceleratorsApple Silicon
HanzoNode
Decentralized AI Compute
hanzod is the node that runs hanzo.network. It joins the fabric over a peer-to-peer transport, serves models to whatever the cluster routes to it, and settles the work on Lux. One Rust binary, started with hanzo fabric up.
Why Run a Hanzo Node
One Rust binary, started with hanzo fabric up.
The hardware you already have
01 / 06
Consumer cards, datacenter accelerators, or Apple Silicon. The node reports what it can serve and the cluster routes to it on that basis.
hanzod01 / 06$hanzo fabric up
reports what it can serve
- Consumer cards
- datacenter accelerators
- Apple Silicon
the cluster routes to it on that basis
Paid out of the block reward
02 / 06
A tenth of the block reward is set aside for AI compute. A provider's share is scaled by the level of the work — embeddings and small models, chat-sized models, large and multimodal, training, or specialized compute like proofs.
hanzod02 / 06$hanzo fabric up
A provider's share is scaled by the level of the work
- embeddings and small models
- chat-sized models
- large and multimodal
- training
- specialized compute like proofs
The cluster knows who has what
03 / 06
Ask it for models and it lists everything served across the peers your node can see. Ask it to route and it names who would answer; ask for placement and it says where a model nobody serves should be loaded.
hanzod03 / 06$hanzo fabric up
modelseverything served across the peers your node can seeroutenames who would answerplacementwhere a model nobody serves should be loaded
Attested, then trusted
04 / 06
A node's confidential-compute tier is attested rather than declared, and its share is scaled by that tier and by a trust score. Claiming a capability is not the same as proving one.
hanzod04 / 06$hanzo fabric up
- declared
- attested
Claiming a capability is not the same as proving one.
Three commands, no orchestration
05 / 06
up starts it on the active network, join switches network and starts it, stop signals the process this CLI started by its recorded pid — and never anything else that happens to share the name.
hanzod05 / 06$hanzo fabric up
up starts it on the active network
hanzodhanzodnever anything else that happens to share the name
Settled on Lux
06 / 06
Work is ordered on the Hanzo L2 and committed to Lux as a Merkle root, so what a node did is anchored where it cannot quietly be rewritten later.
hanzod06 / 06$hanzo fabric up
ordered on the Hanzo L2
committed to Lux as a Merkle root
Put a machine on the fabric
Install the CLI, pick a network, run hanzo fabric up. The node binary is resolved from your PATH or an env var — the CLI runs it and never builds it for you.
Paste and ship
Paste this into any agent. It reads the skill manifest and calls Node from there.
Read https://hanzo.ai/skill.md and use Hanzo Node in my project. Start with: hanzo node list