Requests and CPU-time β model the Workers Paid bill (no egress fees).
Workers meter actual CPU-ms and requests with zero bandwidth fees, so edge apps stay cheap at scale. Compare the wall-clock GB-second model on the serverless function cost calculator.
The Cloudflare Workers Cost Calculator estimates your monthly Workers Paid bill from two inputs: the number of requests per month (in millions) and the average CPU time per request (in milliseconds). The Workers Paid plan starts at a fixed base fee that includes a set allotment of requests and CPU time each month; beyond those included amounts, you pay per additional million requests and per additional million milliseconds of CPU time. Because Workers bills on actual compute time rather than wall-clock duration, the calculator multiplies your traffic volume by your per-request CPU cost to project where you land relative to the included quotas and what any overage adds.
The practical value is seeing which driver dominates your bill before you commit. At high request counts with light logic, the request count usually governs cost; but a function doing heavy work per call can make CPU time the larger line item even at modest traffic. The key trade-off to watch is that trimming CPU milliseconds per request β through caching, lighter dependencies, or simpler logic β often lowers your bill more than reducing requests, since idle time waiting on network or storage generally does not count against CPU billing. Run realistic numbers rather than peak spikes to avoid overestimating.
The Paid plan is $5/month and includes 10M requests plus 30M CPU-milliseconds. Beyond that it is $0.30 per million requests and $0.02 per million CPU-ms. Crucially it bills CPU time, not wall-clock, so time spent waiting on I/O is free.
No egress fees and CPU-time billing. Most serverless platforms charge for GB-seconds of wall-clock duration plus bandwidth; Workers charge only for CPU actually used and never for data transfer, so I/O-bound and edge workloads cost a fraction.
Cost at different volumes:
| Volume | Monthly | Yearly |
|---|