How a CDN works
Your video lives on an origin server. When a viewer requests a segment, the request is routed to the nearest CDN location, called an edge. If the edge already holds that segment it sends it immediately; if not, it fetches it from the origin, sends it and keeps a copy for the next viewer.
Popular content is therefore served almost entirely from the edge, close to viewers, without touching the origin.
Why streaming needs a CDN
- Speed: shorter distances mean faster starts and less buffering
- Scale: thousands of edge servers absorb an audience no single origin could
- Resilience: traffic is routed around failed or congested locations
- Protection: the CDN shields the origin and can enforce signed URLs and geo rules
How CDN usage is priced
Delivery is billed by the volume of data transferred, per gigabyte or terabyte, often with lower rates at higher volumes and different rates per region. Rates vary widely between providers.
Volume is viewing hours multiplied by average bitrate. As a rule of thumb, an hour at 5 Mbps is about 2.25 GB, so 100,000 viewing hours is about 225 TB at that bitrate, or 100 TB at an average nearer 2.2 Mbps.
Reducing delivery cost
- Tune the bitrate ladder so viewers are not sent more than their screens can show
- Use a more efficient codec for your highest-volume content
- Cap quality on mobile data by default
- Compare the platform's included allowance and overage rate, not just its base price
Last reviewed