Plans, billing & usage
Understand token limits and "AI requests are blocked"
Where AI usage is measured, the two messages you will see when a cap is hit, and how an admin raises it.
Where usage is counted
Everything an agent or Ellum Assist generates costs tokens. Settings, then Usage & Limits shows the organisation's monthly and daily meters, and lets an admin set a separate limit per member and per agent. Setting one to 0 pauses that person or agent without touching anyone else.
The two messages
"Over the limit — AI requests are blocked." means a meter is full. It clears when the period resets or the cap is raised. "Paused by an administrator — AI requests are blocked until the cap is raised or removed." means someone set the limit to 0 on purpose.
Raise the cap
- 01Open Settings, then Usage & Limits.
- 02Find the member or agent, or the organisation meter.
- 03Raise the number or clear it to inherit the plan limit.
- 04If the plan itself is the ceiling, compare limits on the pricing page.
Bring your own key
Settings, then AI Providers lets you add your own Google, OpenAI or Anthropic keys. Keys are stored encrypted and the raw value is never shown again after saving. Each key has a daily request cap that resets at midnight UTC, and you can order which provider is tried first.
Related