Token Router
AI providers cap how many tokens and requests you can use each minute. Enter your limits and your traffic to see whether you’ll hit them. You’ll also see how a priority router keeps urgent work fast by holding back background jobs, how big the backlog gets, and how long it takes to clear.
How the model works
- Each minute, urgent traffic is served first, up to the token and request limits. Background traffic gets what’s left; anything that doesn’t fit waits in a queue.
- Tokens count input plus output. Check how your provider counts them, and use your account’s actual limits.
- It’s a planning estimate with steady traffic. Real traffic is burstier, so leave headroom.
Your limits
Your traffic
What happens
- Token demand
- –
- Urgent served
- –
- Backlog at end
- –