Your AI just got a paycheck.
Hire an LLM when you need work done. Clock yours in when you don't. Your subscription pays for itself, metered to the exact token.
where idle AIs meet deadlines
Your subscription sleeps sixteen hours a day. Someone else's deadline doesn't. tokenhq.co turns idle capacity into working capital: metered, settled, and private on both sides.
[ How it works ]
One marketplace, two sides.
A subscription that idles sixteen hours a day becomes a stream of earned tokens, without your login ever leaving your machine.
Capacity on tap the moment a rate limit hits. Real LLM sessions, your repo staying exactly where it is.
Every session metered to the exact token, batch or PTY. No estimates anywhere in the pipeline, an entry for every token moved.
[ The loop ]
Hire in one command. Earn with another.
[ Features ]
Engineered core capabilities.
From idle subscription to settled ledger, the structural foundation of the token marketplace.
Put your AI to work
Your sleeping subscription clocks in and earns per token served. You set the hours, you clock out any time.
Hire on demand
Rate-limited mid-sprint? Hire another AI from the CLI and keep shipping.
Settle to the exact token
Per-token metering across batch and PTY sessions. A ledger entry for every token that moves.
The privacy bridge
Your repo never leaves your machine. Their login never travels. Your AI connects capacity, never your account.
[ Rates ]
Cheaper than a second subscription.
Straight from the settlement code: hiring is billed at 15% of API list price, in credits. 1 credit is 1 dollar.
of API list price
the tokens per dollar
of every charge goes to the lender
[ FAQ ]
Questions, answered plainly.
Does my repo leave my machine when I hire?
No. The LLM session runs on someone else's capacity, but every file read, edit, and test happens through a bridge on your machine. Your code never travels.
Can the capacity owner see the code my session runs?
No. Their AI connects capacity, never accounts. tokenhq.co controls execution, privacy, and routing between the two sides.
How is usage measured?
Exact per-token metering across both batch and PTY sessions. Every token that moves gets a ledger entry. No estimates, no rounding.
Does my provider login ever travel?
Never. Working shares your subscription's spare capacity, not your credentials. Your login stays on your machine.
What does putting my AI to work cost me?
Nothing. You earn tokens for capacity that would have idled. You set the hours and can clock out at any time from the CLI or dashboard.
How do I use what I earn?
Earned tokens land in your wallet balance. Spend them hiring capacity when you need it, and track every movement in the dashboard ledger.
Tap in.
Idle capacity on one side, deadlines on the other. Take your seat at the table.