The supply side of the agent economy

Get paid when
AI gets to work.

Agents need compute, models, and services to act in the real world. Supply one — and earn every time they use it.

For device & node operators

Run compute

Host inference on your phone, GPU, or server. Earn USDC every time an agent routes a call through your node.

  • Phones, consumer GPUs, and data-center hardware all supported
  • Pricing set by node — Ghola matches agents by price, latency, and reputation
  • 85% of per-call revenue to the operator, 15% protocol fee
Host a node
For AI creators

Publish a model

Fine-tune and ship a model under your name. Agents find it, call it, and pay per inference — you earn per call.

  • Fine-tune from your data or bring a model you own
  • Your model gets an on-chain identity and reputation score
  • 85% revenue share, hourly USDC settlement
Publish a model
For API operators & businesses

Sell a service

Register your API as a headless merchant. Agents discover you, verify you, and pay per call — no checkout, no accounts.

  • Per-request, per-minute, or per-token pricing in USDC
  • We handle metering, auth, and settlement — you get a revenue dashboard
  • 3% platform fee, hourly settlement
Register a service

Same rules. Every side.

Whatever you supply, you get the same economic primitives.

USDC settlement

Every transaction settles in USDC on Solana, per-call, hourly batched.

On-chain reputation

Every call builds your score. No gatekeepers can take it from you.

Open protocol

Built on SAID. Anyone can plug in, anyone can fork, no walled gardens.

Not sure where to start?

Start with what you already have. Phone in your pocket? Run compute. A model you've fine-tuned? Publish it. An API? Register it.

See how the full flywheel works