Skip to main content
Each Azura model has a rolling usage window. The specific limits and reset windows can change over time, so this article explains how the system works rather than listing numbers.

​
Reading the context bar

The context bar above the Omnibox shows:
  • Used / limit for the current model.
  • A progress bar that changes color as you approach the limit.
  • “resets ” — when your usage resets.

​
When you hit a limit

Sending is blocked with a message naming the limit type and the exact reset time, such as “Token limit reached. Limit resets at .” Your options:
  1. Wait until the reset time shown.
  2. Switch to a different model. Each model’s quota runs independently, so a model you have not used still has its full quota available.
This is expected usage metering, tracked per model, per account — not a bug.

​
Check your usage

Use the /tokens slash command in the Omnibox to see your current usage for the period.

​
Long conversations

In very long conversations, Azura automatically summarizes older messages to stay within limits. The most recent messages are always kept in full; older context is condensed into a summary. From your perspective: in very long chats, Azura may lose track of details from much earlier in the conversation. If that happens, start a new conversation and bring the important details forward.