Usage & Quota
When you use Ora through the managed Aethrix profile, your usage counts against the allowance included with your license. Ora shows where you stand directly in the panel, so a limit never takes you by surprise.
Credits
Managed AI usage is measured in credits. Each request draws credits based on the work it does — the model behind it, the size of the input and generated output, and how much of it is served from cache — at the rates in effect when the request happens. Credits are the only usage figure shown inside Studio. The AI & Usage page of your account portal goes one step further and lists the input and output token counts behind each recent request, so you can see why a particular turn cost what it did; hovering a row breaks the prompt into the parts that are priced differently. No surface anywhere states a money value for a credit — a credit is a unit of consumption. And each request is priced at the moment it happens, so if rates change, your past usage keeps the value it had when it happened.
The allowance works like a monthly plan: you get a fixed number of credits per period, and once they are spent, further requests are refused until the window resets. Fractional usage adds up — a short turn may cost a fraction of a credit, displayed with two decimals.
The quota chip
At the bottom of the Ora panel — left of the agent and model pickers — a small chip shows your remaining headroom:
- A colored dot: green while you are comfortably inside the allowance, amber from 80% of the tighter window (day, week or month) upward, red when a limit is reached.
- A percentage of the tighter window that is still available (for example
94%). On an organization's own provider, a window your administrators left uncapped shows the amount consumed instead of a percentage.
Click the chip to open the detail flyout:
- Today (only when your plan sets a daily ceiling), This week (only when it sets a weekly cap) and This month, each with a progress bar, the exact
used / limitcounters in credits, and the reset time — daily windows reset at 00:00 UTC, weekly windows on Monday 00:00 UTC, monthly windows on the 1st of the month (UTC). - Manage in account portal — opens the AI & Usage page of your account portal, which repeats the Usage today and Usage this month figures as cards, adds a Usage this week card when your plan sets a weekly cap (a plan with a lifetime grant shows a Lifetime allowance card instead of a monthly one), adds a Usage under your licence card when your account carries a license whose pool has drawn credits this month — that total follows the license rather than the person, so an administrator reassigning the license mid-period does not start it over — then adds a daily credits chart, a per-model breakdown, and a Recent requests table giving each request's model, Input and Output token counts and the credits it drew. The Input figure is the whole prompt; hovering it breaks out how much arrived uncached and how much was read from or written to cache.
The chip refreshes automatically after each conversation turn, when you switch models, and when you open the flyout.
What the header means
The flyout header tells you which allowance is being consumed:
| Header | When you see it |
|---|---|
| Pro included | The managed Aethrix backend with a Professional license — AI usage is part of what the license includes. |
| Enterprise included | The managed Aethrix backend with an Enterprise license. |
| <name> — Provided by your organization | Your organization configured its own LLM provider (for example "InternalGPT"). Usage counts against the per-user quota your organization's admins set. |
| (no chip) | A profile you configured yourself with your own API key (BYOK). Aethrix meters nothing there — your provider bills you directly. |
Included allowances
Default allowances per license tier, in credits (your deployment may configure different values, and administrators can raise them per user, per organization or per tier):
| Tier | Daily | Weekly | Monthly | Lifetime |
|---|---|---|---|---|
| Community | — | — | — | — |
| Professional | — | 200 credits | 500 credits | — |
| Enterprise | — | 400 credits per member | 1,000 credits per member (pooled: members share a pool of 1,000 × seat count) | — |
Community accounts have no Aethrix-managed AI: they either bring their own provider key (BYOK), or earn the 14-day Professional trial through the tutorial — and a trial account is metered as Professional, so it draws on the allowances above.
The weekly figure is a ration set at 40% of the monthly allowance, so the month's credits are spread across the month instead of being spent in its first days — it is the shipped secondary window, not the daily one. A daily ceiling is still available but ships unset: an administrator can add one, or change or remove the weekly cap, per user, per organization or per tier from the backend without a deployment — enterprise administrators manage their own members' windows this way.
When a window's limit is reached, further requests are refused until that window resets; the chip turns red and shows Limit reached.
Enterprise organizations
If your organization routes Ora through its own LLM provider, the chip header shows the provider name your administrators chose, followed by Provided by your organization. The numbers come from the quota your organization set for you; depending on how your administrators configured enforcement, reaching the limit either blocks further requests or only shows a warning. There is no account-portal link on this chip — per-user detail is a matter for your organization's administrators.
Organizations using the managed backend share an org pool: by default each member brings 1,000 monthly credits into the pool. Enterprise admins see pooled org consumption on both the portal Dashboard and the Analytics page, counted over the recent days the gateway reports rather than a calendar month, and every member can see their own usage on the AI & Usage page of the enterprise portal.
Notes
- The chip only appears while a managed or organization profile is the active model; switching to a BYOK profile hides it.
- The same credit numbers appear in the chip in Ora, the AI credit usage card on Studio's My Account page, the AI Credits Used card on your portal dashboard and the Subscription page, and the AI & Usage page of your account portal.
- The AI credit balance granted by the trial tutorial is a separate ledger that is not debited by gateway metering yet. It no longer appears as a credit meter in the portal — every portal credit figure comes from the gateway.