The problem with most DDoS pricing
DDoS protection is usually sold one of two ways. A flat fee that ignores how big your attacks are, so quiet customers subsidise noisy ones. Or a bill that grows with the traffic the provider filtered, so you pay more the harder you get hit.
Templass does neither. The invoice tracks the traffic that survives scrubbing and reaches your origin. The flood we drop at the edge is not a line item.
What counts and what does not
- Counted: traffic forwarded to your origin after mitigation, at the 95th percentile of clean samples in the billing period.
- Not counted: attack volume dropped at the edge, a short spike that falls in the discarded top 5 percent, and your ISP uplink or GRE outer overhead.
Edge capacity in Frankfurt, in the over 1 Tbps class, is the platform, not something you buy. Your quote is about clean Mbps, not about buying a terabit.
How the 95th percentile works
Templass samples your clean rate through the billing period. The highest 5 percent of samples are thrown away. The highest sample left is the 95th percentile, and that is the number you are billed on.
In practice, a few short bursts do not set your month. A clean peak that lasts most of the month does. The pricing page has a small chart of the idea.
The indicative rates
Rates are whole-volume. The entire measured clean Mbps is charged at the single tier the measurement falls into.
- 0–299 Mbps: €0.45 per Mbps
- 300–509 Mbps: €0.41 per Mbps
- 510–999 Mbps: €0.39 per Mbps
- 1000+ Mbps: €0.33 per Mbps
A €39 per month floor applies on quiet months. You pay the higher of the tiered charge and €39, never both.
Working out your peak clean Mbps
The quote form asks for one number: the clean traffic you expect at the origin, not the size of the attacks you have seen. A good starting point is your traffic graph on a busy but normal day, ignoring the spikes caused by floods.
Because billing uses the 95th percentile, short evening peaks matter less than you might expect. If you are not sure, give a range in the notes field and say what kind of service it is. A game server, a hosting node and an API all behave differently, and the quote can reflect that.
The other inputs that shape the quote are the handoff, GRE or Cross-Connect, and anything in the notes such as prefixes, ports and where the origin lives.
From quote to invoice
- Request a quote with your use case, peak clean Mbps, and GRE or Cross-Connect.
- Get a written quote with the rate, any commit, the handoff and the term.
- Accept it in writing. Email is enough. Provisioning starts from there.
- Pay the invoice using the method it names. The Terms list Stripe, PayPal and EU or SEPA transfer.
Need more or less clean bandwidth later? That is a new quote, not a plan switch. Read DDoS protection over GRE or DDoS protection in Frankfurt for the handoff side.