Operators
Demand for the machines you already run.
Tenzro brings agent and app demand to operators of every size, from a single server to a data center. You set the price for each model and unit of work. Each call is paid against a receipt your machine signs, and every delegated agent that pays you acts under limits set by a named responsible party.
Why connect
What connecting gives you.
Demand, routed to you
Calls arrive through one API and are routed by price, latency and reputation. A failed or silent provider is skipped within seconds and the call moves on.Paid per call, at your rate
You publish a rate card per model and per unit of work, and the network sets no price. Each call settles against a receipt your machine signs, capped by what was actually sent and received.Spot work and standby fees
Idle capacity takes spot work, and capacity you hold in reserve for a buyer earns a standby fee.Paid for being available
An availability reward pays machines that stay ready, checked by challenges they cannot predict.Private models, safely served
Private and independent model providers serve on your GPUs without exposing the model. Weights stay sealed and are released only to an environment whose attestation verifies.Channels for frequent work
Payment channels carry many small calls without a transaction each; funded escrow pays for capacity you hold for a session.Accountable counterparties
Every delegated agent that pays you acts under limits set by a named responsible party. A revoked agent stops paying at once, everywhere.Your policy, enforced
Publish what you serve and where. A call pinned to a jurisdiction gets a signed jurisdiction receipt or a refusal.Every machine can take part
Each machine carries a visible security level and hardware type. Stronger hardware earns a capped bonus, and older hardware keeps its full base reward.
Every scale
Built for the hardware you already have.
- At home
One machine with a GPU
Machines of every size can serve: a workstation, a GPU server, a data center. The machine's TPM or your passkey holds the keys. - Homelab
A few machines, one cluster
Pool machines on your own network into one node. Serve larger models across them, or split roles between them. - Data center
Independent data centers of any size
One operator enrols whole fleets and racks, including rack-scale GPU systems, with a hardware identity per host and a device certificate per GPU. The node runs vLLM, SGLang or TensorRT-LLM in pinned environments beside your own tooling. - Neo-clouds
Distribute capacity and models
Reach developers, researchers and agents directly, with rental billing on the chain's verdict and SLAs you can prove.
Node tiers
One codebase, a build for every machine.
Each machine runs the release its hardware can serve and refuses the roles it cannot.
- Edge
Small and embedded devices
Single-board computers, Jetson-class modules, mini PCs, NAS boxes and gateways. They hold identity, make payments, run agents and storage, and pass inference on to GPU providers. - Provider
CPU servers and workstations
Every role a CPU can serve, including the models that fit it. - Provider
From one GPU to a data center
Inference, compute rental and confidential computing, with models too large for one machine served in pipeline shards across your own machines, and mixture-of-experts models across holders on the network. - Validator
Servers with hardware-held keys
Order transactions and certify checkpoints with keys held in a TPM 2.0, a Secure Enclave or a delegated key holder, against a bonded stake.
Roles
Ways to provide resources.
One machine can take on several roles from one account, with a bond for each. Pick the ones your hardware and your connection suit.
Not running hardware? You can also operate as a technical support or go-to-market partner.
How it works
From install to first payment.
- 01
Install
Download the node and run tenzro. It checks your machine, detects your accelerators and sizes itself to the hardware.
- 02
Create your identity
Your operator identity is rooted in the machine's TPM or Secure Enclave, or in your passkey. No key files to guard.
- 03
Stake and choose roles
Bond stake for each role the machine takes on. Publish a signed policy for what you serve and at what price.
- 04
Serve and get paid
Each call is metered on your node and paid per call against a receipt your machine signs.
Accountability
Accountable to what you promised.
Operators publish signed policies. Slashing applies to breaches of your own published policy and of the SLAs you sign, decided by validators holding two thirds of stake, so rentals are billed on the chain's verdict rather than anyone's word.
Help getting started
Onboarding for every background.
The Foundation publishes the node, the guides and the tooling, and partners help operators of every size and experience level get their hardware on the network.
Provide compute
Ready to operate?
Pick the roles that fit your hardware and set your own prices. You are paid per completed unit of work, and what you earn follows the paid work you serve.