UsePod

UsePod is an optional marketplace for AI model capacity. Use it as a prepaid model provider for an agent. Advanced operators can also offer their own local compute on it.

Set up UsePod for an agent

1. Open the agent’s model or provider settings and choose UsePod.
2. Complete the guided registration.
3. Fund the deposit address shown with only the amount you intend to use.
4. Refresh the model list and run the short connection test.
5. Pick a spending preset or set price limits before normal use.

HivemindOS stores the setup securely and shows your balance and route details when UsePod reports them.

Billing

Two billing options may be available:

• Your UsePod setup uses your own UsePod account or prepaid balance.
• HivemindOS-managed billing uses eligible HivemindOS service credits.

Before any paid use, the app shows which option the agent uses, its limits, and the current price or fee. Trust what the app shows, not an old percentage in documentation or a local setting.

Control spending

• Cheapest uses strict price limits.
• Balanced allows a wider range of routes with guardrails.
• Fast allows pricier routes when speed matters.
• No cap removes the price ceiling. Use it deliberately.

These limits are separate from wallet budgets, company budgets, and managed credits. A request must pass every limit that applies.

Offer your own compute (advanced)

Hosting is a separate decision from using UsePod. It can expose a local model server to marketplace work and may need identity keys, a funded operator bond, and network setup.

1. Open the machine in Fleet and start its compute-hosting action.
2. HivemindOS checks for a compatible local model server, finds its models, and prepares the worker. It will not go live if a check fails.
3. Review the advertised models, concurrency, network exposure, earnings policy, and stop controls.
4. Choose Go live.

Connecting UsePod for an agent never turns on hosting by itself.

Privacy and availability

Requests sent to UsePod leave your machine. Keep sensitive work on a local model.

Available models and prices change. If a model is no longer available, HivemindOS reports it rather than silently switching to a provider with different billing or privacy, unless the route you chose includes that fallback.

For protocol details, see the UsePod documentation.

Related

• Wallets and spending controls
• Integrations