Pathwise
Technology
Idle Macs. Cheap, private tokens.
Connect idle Apple Silicon and earn USD or USDC. Or use the Pathwise API for private open-model inference — OpenAI-compatible, at a fraction of what the AI labs charge.
Payments on
BASE · Etherlink · USD / USDC
Private by design
Prompts stay off the operator's screen and off-chain. Data is encrypted end-to-end and never exposed publicly.
Run on Apple Silicon
Turn the Mac you already own into an inference node. No specialized infrastructure.
Earn with every job
Get paid in USD or USDC on Base and Etherlink, plus Path points toward a future $PATH snapshot.
Built for agents
Call the Pathwise API for cheap AI tokens. OpenAI-compatible. Agents pay over x402 — no checkout, no keys.
The core proposition
Idle Macs on one side. Cheap tokens on the other.
Pathwise turns idle Apple Silicon into a private inference network. Developers and agents call an OpenAI-compatible API and buy tokens at a fraction of lab prices. Jobs run locally on connected Macs.
Mac operators earn USD or USDC for every completed job, plus Path points. No new hardware. No cloud lock-in. No prompts on your screen.
How it works
One network. Three steps.
01
Connect
Connect your Mac and choose when you want to contribute compute.
02
Compute
Encrypted inference jobs are routed to your machine and executed locally.
03
Earn
Receive USD or USDC for completed work, plus Path points.
For Mac operators
Turn idle compute into income.
Your Mac doesn’t have to sit idle. Contribute unused Apple Silicon to inference jobs and get paid for the work your machine performs.
You stay in control. Pause anytime. Monitor your activity. See exactly what you’re earning.
For developers
Inference without the infrastructure bill.
Build with an OpenAI-compatible API and access distributed Apple Silicon through Pathwise. Send a request. We handle routing, execution, and settlement.
- OpenAI-compatible API
- Distributed Apple Silicon inference
- Encrypted job routing
- USD / USDC settlement
- Base + Etherlink
- x402-native payments
For AI agents
Agents that can pay for compute.
With x402, an agent can request inference and pay for it in the request flow. No human checkout. No prepaid credits. No API key management.
01
Request
02
Pay
03
Compute
04
Receive
Two ways in
Connect a Mac. Or just call the API.
Operators earn from idle Apple Silicon. Developers and agents buy private inference at a fraction of lab prices.