Networking
Latency Estimator
Estimate round‑trip time (RTT) from distance, or model request latency from RTT + DNS/TLS handshakes. Local only.
Inputs
Use the distance mode for a theoretical lower bound. Use the RTT model for a practical “how long will this request take?” estimate.
Light travels slower in glass than in vacuum (~2/3c). Real routes also take detours.
Result
Propagation-only estimate (lower bound).
Estimated RTT
10 ms
Practical RTTs are usually higher than propagation-only due to routing, queuing, and processing.
One-way (min)
5 ms
RTT (propagation)
10 ms
Practical lower bound
13 ms
What this tool estimates
Latency is mostly about round trips. Even tiny payloads can feel slow when RTT is high.
- Distance mode estimates propagation delay (speed-of-light limit).
- RTT model estimates request time from how many RTT steps occur (DNS, TCP, TLS, request/response).
Tips to reduce latency
- Reuse connections (keep‑alive) so you don’t pay TCP/TLS every request.
- Use a CDN/edge to move content closer to users.
- Reduce request count and avoid redirect chains.
- Enable HTTP/2 or HTTP/3 to cut head‑of‑line blocking and handshake costs.
Common questions
Why is RTT higher than the distance minimum? Routes are not straight lines and packets traverse many devices; congestion and buffering add more delay.
Is bandwidth the same as latency? No. Bandwidth is “how much per second”; latency is “how long it takes to get a response.”
Assumptions & limits
- Distance mode ignores routing/queuing and only models propagation.
- RTT model assumes RTTs happen sequentially (real browsers can parallelize some work).
- Large downloads/uploads depend on throughput too; use the bandwidth calculator for that part.