Usage & Limits

This is a rate-limit, not a bill

First, the important part: the AI usage system is about request limits, not money. Nothing here charges you, and there is no per-server billing. It exists so the bot stays fair and stays up.

Here's why it has to exist. exceed talks to Google Gemini, and Gemini's free tier has hard daily caps on how many requests the whole bot can make — and that cap is shared across every server exceed is in. If one large, busy server fired AI commands non-stop, it would burn through the shared daily ceiling and every other server's AI would stop working.

To prevent that, exceed gives each server its own monthly request and token allowance, plus per-user cooldowns. Think of it as a courtesy budget: your server gets a fair slice so it can't accidentally (or deliberately) starve everyone else.

The three controls

All of this lives on the AI Usage dashboard page (titled "AI Usage & Budget"). It applies to every AI feature on the server at once — summaries, persona chat, the embed generator, mod copilot, smart help, the setup assistant, and image automod all draw from the same pool.

1. Monthly call cap

The maximum number of Gemini API calls your server can make in a calendar month. The default is 2,000 calls. Set it to 0 for unlimited.

2. Monthly token cap

The maximum total tokens (input + output combined) your server can use in a month. The default is 5,000,000 tokens. Set it to 0 for unlimited.

3. Per-user cooldown

How long a single member must wait between AI requests, in seconds. The default is 8 seconds. This stops one person from spamming AI commands and chewing through the whole server's allowance in a minute. Set it to 0 to disable cooldowns.

The "Usage this month" meter

The top of the AI Usage page shows a live meter of how much you've used this month:

• API calls — used vs. your call cap.
• Tokens — used vs. your token cap.

The bars turn amber at 80% and red at 100% so you can see at a glance how close you are. The meter also appears on the Persona Chat and Image AutoMod pages, since those are the features most likely to consume budget.

Allowances reset automatically at the start of each calendar month (UTC). Usage is tracked per period (YYYY-MM), so a new month starts you fresh.

The master switch

The page also has an Enable AI features toggle. This is the single kill switch for all AI on the server:

• On — AI features work, subject to the caps above.
• Off — every AI command immediately stops and returns a short "AI features are disabled on this server" message. Use this if you want to pause AI entirely.

What happens when you hit a cap

When your server reaches its monthly call cap or token cap, AI commands reply with a friendly message — "This server has reached its monthly AI usage limit. It resets at the start of next month." — instead of failing silently. Features that run in the background (like image automod or the ticket assistant) simply pause until the budget resets. Nothing breaks; AI just takes a rest until the next period.

Quick reference

Manage all of these on the AI Usage page for your server.