Frequently asked
What does the Free plan include?
200 AI requests per month with 500K input / 150K output tokens, one registered installation, local models, and baseline cloud models. No card required.
How is usage billed?
Requests consume your plan allowance first. If your plan enables overage and you're within your spend limit, excess usage is priced server-side by versioned rules per model and category — you can reconcile every record in the dashboard.
Does local inference cost anything?
No. Records marked mode=LOCAL (running entirely on your hardware) are metered for your analytics but never billed. The classification is enforced server-side.
Can I set a hard spend cap?
Yes — Pro and Enterprise subscriptions support a spend limit. Once reached, billable overage stops automatically; allowance usage continues.
How does the desktop app authenticate?
The app never stores your password. Signing in registers a revocable installation and issues a short-lived access token plus a rotating refresh token, stored in your OS credential store.
What happens if I revoke an installation?
Its tokens stop working on the next request — the app receives INSTALLATION_REVOKED and returns to signed-out state. You can re-register at any time within your plan's installation limit.
How do organizations work?
Enterprise accounts create organizations with Owner/Admin/Member roles, invitations, centralized billing, org-level usage rollups, and entitlements resolved per organization.
Is my code sent to the billing platform?
No. Usage records contain measurements (tokens, tool calls, timing) — never prompts or source code. That boundary is contractual, not just conventional.