Skip to main content
Every paid plan has three limits. Each is a share of the month’s usage, shown in percent: On Go, one 5-hour window can use up to 20% of the month’s usage and one week up to 50%, so a long session can’t empty the month in a day.

The three windows

  • 5 hours. The window starts with your first request and resets 5 hours later. The next one starts with your next request after that.
  • Week. Runs in 7-day steps from the day your billing cycle starts.
  • Month. Your billing cycle. The month’s usage renews when it does, and usage you don’t spend doesn’t carry over.

Past your limits

At 100% of a limit, new requests pause until that limit resets. With extra usage on, they carry on using purchased credit instead. Past 100% of your plan’s monthly usage, new requests stop: they return 402 allowance_exhausted. The message says how much of the month you have used, when your usage renews and which plan has more. In-flight requests finish: nothing is cut off mid-stream, and your subscription is never billed extra for it. A 402 carries no retry-after, because waiting a few seconds doesn’t clear it. Every paid plan also has a 5-hour and a weekly usage limit. Past one, requests return 402 five_hour_limit_reached or 402 weekly_limit_reached with the time the limit resets (resets_at). Three ways to keep going:
  • Turn on extra usage on the billing page. Past your monthly usage or a usage limit, requests carry on using purchased credit, with your plan’s model access, when you have enough.
  • Use purchased credit in PAYG mode. Add credit on the billing page, then turn on PAYG mode for your key or send the x-sideren-payment-mode: payg header. The setup helper’s npx sideren setup --payment-mode payg stamps it into the agents it configures.
  • Upgrade. The higher plan’s usage and limits apply right away. Downgrades take effect at renewal.
On Free, an account still holding a trial balance from before gets 402 credit_used_up when it is used. Purchased credit or a subscription keeps you going.

Where to see them

The dashboard shows a bar for each limit and when it resets. Every response carries the same figures as headers (Headers): The headers also ride on 402 and 429 answers, so a client can read what is left when a request is refused. Alert on the percent before it reaches 100.