Roles and limits

The monthly limit#

Every server has one monthly limit covering everything Xertron does in it. It resets on your signup anniversary rather than on the first of the month.

Until you split it, everyone in the server draws on the same pool, which means one enthusiastic member can spend a month's worth in an afternoon. Splitting it is the single most useful thing to do on the dashboard's Roles & limits screen.

Any member can run /usage at any time, on any plan, to see how much of the limit is gone, how long until it resets, and their own role's share if one applies. It posts publicly in the channel and changes nothing.

How a role is chosen#

A member's limit and feature access come from the first role in the list that applies to them. The rest are ignored. Roles do not stack, and holding more of them does not add anything up.

You set that order yourself by dragging the rows, so priority is not Discord's own role hierarchy. @everyone sits at the bottom as the baseline for anyone no override matches.

Splitting the limit#

Each override takes a percentage of the monthly limit. The percentages partition the limit between your roles, so the screen tracks what is allocated and tells you when you are fully allocated or over.

A share is a slice of the whole limit, not an allowance per person. The bar for a role tracks what everyone holding it has spent between them, which is exactly what decides when its members start getting lighter answers.

If there is not enough unallocated headroom left to give a new override its minimum, the screen says so and the add button waits. Lower another role first.

Opening premium features#

Each role override carries its own grants, separately from its share. These are the three that cost noticeably more per message.

• Pro model. Whether !pro and the Pro option in /settings do anything for them.
• Extended thinking. Whether they can reach the deeper reasoning levels.
• Image generation. Whether /imagine works for them at all.

These sit underneath the server-wide toggles rather than beside them. A feature turned off for the server is off for everyone regardless of role, and a role grant only decides who reaches something the server already allows.

This is also why a control can be missing from a member's /settings. Each option only appears when it would actually do something for them, so an absent control means their plan, the server's settings or their role does not currently grant it.

Running out#

As a share is used up, Xertron steps down in capability rather than stopping. The Pro model gives way to flash, deeper thinking gives way to standard, and answers keep arriving. It tells people plainly when a reply came out lighter for that reason.

A server can also carry a top-up balance that gets spent once the monthly limit is gone. There is currently no way to add funds to it from the dashboard or anywhere else in the product, so treat the monthly limit as the real ceiling for now.

Where to watch it

The dashboard's Usage & billing screen is read-only and shows the cycle meter, a day-by-day spend chart broken down by request kind, by model and by role, a wallet card per role showing what it has spent of its own slice, and a filterable log of individual requests.

If you would rather cap spending than watch it, eco mode on the Overview screen holds everyone to the fast model with no extended thinking, whatever they ask for.

Blocking one member#

Blocking is separate from all of the above. It stops one specific member using Xertron in the server at all, rather than giving them a smaller share.

Add them from the member list on the same screen, with a reason, so the next admin reading it knows why. Only an admin can lift it.

For what the plans themselves include before any of this applies, see Plans.