Skip to main content
Whether you use MiniMax Code or other AI tools connected with your Subscription Key, it all counts toward the same M Plan usage limits.

​
Usage resets

Usage resets automatically over two time windows. No manual action is needed, and you do not have to wait until your renewal date.
  • Non-video models require remaining usage in both windows; video models are subject only to the weekly window.
  • The windows reset independently. A 5-hour reset does not reset your weekly usage.
  • Each window resets when it ends, regardless of your usage during that window. For example, if your first use is at 10:00, that 5-hour window ends at 15:00; even if you reach your usage limit at 12:00, it resets at 15:00.
  • Unused usage does not accumulate in the next window or carry over to the next billing period.

​
Check your usage

Your usage history helps you understand how you work and whether your current tier fits. On the Plan usage page, you can see your current plan’s usage and detailed usage analytics.
With MiniMax CLI, run mmx quota to check your usage directly from the command line.

​
When you reach your usage limit

  • Upgrade your plan: If you often reach your usage limit, you can upgrade to a higher tier. See Upgrades and cancellation.
  • Wait for a reset: When you reach your 5-hour or weekly limit, your usage resets automatically when that window ends.
  • Use purchased credit packs: For models and capabilities supported by credit packs, the system uses your included plan usage first, then deducts any overage from your available Credits.
  • Use pay-as-you-go billing: For resources outside your plan or for production workloads, use a regular pay-as-you-go API Key. Charges are deducted from your Open Platform account balance. This Key is separate from your Subscription Key and cannot be used interchangeably with it.

​
Why am I rate-limited before reaching my usage limit?

Usage limits and request rate limits are separate. Sending too many requests in a short time, or encountering dynamic limits during peak hours, can temporarily block calls even before you reach your usage limit. Reduce your request frequency or concurrency and try again later. A rate limit does not mean you have reached your usage limit.