Skip to main content

Longer period/larger usage quota

Feature Request

With the new usage system, I can send a single message with a pricy model and cannot reply with those same settings where previously I would be able to send many responses.
Because of this, I do wish that instead of having 4hr(?) blocks of quota I could set a longer period where the usage would be scaled to that longer period of time. That would allow me to do the longer threads/more expensive models and not have to worry about being over usage while I still need more use.

Status: Closed1 comment

Log in to comment and vote

Comments1

  • Mark Florkowski

    Team

    Thanks for the feedback. I can definitely see how a daily reset might feel more flexible for some workflows.

    where previously I would be able to send many responses.

    The reason why this system was changed was because this type of usage was very unsustainable. Different messages cost different amounts, and the new system accounts for this directly.


    Usage comes from two buckets: a base bucket that refills every 4 hours and a larger overage bucket that refills monthly. The “overage bucket” exists to handle days of heavier burst usage, while the “base” bucket handles most day-to-day use.

    Rolling replenishment helps allow us to offer higher effective usage overall. If the same allocation were converted into a large daily bucket, many users would consume it all at once, which would increase peak infrastructure costs and likely require lower overall allowances to remain sustainable.

    If you are consistently hitting limits on the Pro tier, it may indicate that you would be better suited with the Premier tier which offers higher usage limits.