Ping Test
Measure real-world HTTP round-trip latency from your browser — not classic ICMP ping, which browsers cannot send. The tool probes Cloudflare, Google, and Quad9 DNS-over-HTTPS endpoints across multiple rounds and reports average, minimum, maximum, and packet-loss-style failure rates. Optionally add your own hostname to compare latency to a specific site via the ToolsLibrary uptime probe. Useful for checking whether slow browsing is local Wi‑Fi, ISP routing, or a distant endpoint — for legitimate troubleshooting only.
How to Use the Ping Test
Run HTTP latency samples against global DNS endpoints and optionally your own host. Review averages, min/max, and loss in the summary table.
- Open Ping Test.
- Leave Optional host blank to test only Cloudflare, Google, and Quad9 DNS endpoints.
- Or enter a domain (example.com) to add a custom HTTPS uptime probe to the target list.
- Set Rounds (1–10; default 4) — higher values average out jitter but take longer.
- Click Ping and watch the live sample log while probes run.
- When finished, read Overall average and per-target Avg, Min, Max, and Loss columns.
- Compare results over time or on a different network to isolate local vs remote latency issues.
HTTP latency vs ICMP ping
Because browsers block raw ICMP, this tool uses HTTPS requests to well-known DNS resolvers. That measures the same kind of encrypted traffic most websites use, making it a practical latency indicator even though it is not identical to ping in a terminal.
Worked example
- High loss to all three DNS targets — likely local Wi‑Fi, VPN, or ISP issue.
- Good DNS latency but slow custom host — problem may be the target server or CDN region, not your connection.
- Run again with 8–10 rounds before calling your ISP to rule out transient jitter.
Limitations
Results vary by device, browser, VPN, and time of day. Custom host probes originate from the ToolsLibrary server, not your browser's network path. This is not a SLA-grade monitoring service.
Acceptable use
Use for personal network troubleshooting and education. Do not automate high-frequency probes against third-party infrastructure without permission.
FAQ
Is this ICMP ping?
No. Web browsers cannot send ICMP echo (ping) packets. This tool measures HTTP/HTTPS round-trip time to public DNS endpoints, which correlates with real browsing latency.
Which targets are tested by default?
Cloudflare DNS, Google DNS, and Quad9 DNS — each via their public DNS-over-HTTPS URLs. These are globally distributed reference points.
What does the Rounds setting do?
Each round sends one request per target. More rounds (up to 10) smooth out jitter and give more reliable averages but take longer.
How do I test my own website?
Enter a hostname in Optional host (for example example.com). It is probed via HTTPS through the ToolsLibrary network API alongside the default DNS targets.
What does loss percent mean here?
It is the percentage of HTTP requests that failed or timed out (about 8 seconds per attempt) within the test — analogous to packet loss in ICMP ping.
Why are my results higher than a desktop ping utility?
HTTP probes include TLS and application overhead, and they run in the browser sandbox. Desktop ICMP ping measures a different layer and may use a different network path.
Can I use this to attack or flood a host?
No. Rounds are capped and requests are rate-limited by browser and server constraints. Use only for reasonable latency checks on hosts you manage or have permission to test.
Does the optional host probe run from my computer?
The custom host check uses the ToolsLibrary server-side uptime endpoint, so its latency reflects the server environment — not your local ISP path to that host.
Why does the test take several seconds?
Multiple targets × multiple rounds run sequentially by round (targets within a round run in parallel). A 4-round default test with three DNS targets performs twelve samples.
What is a good latency number?
Under 50 ms to nearby reference DNS is excellent on broadband; 50–150 ms is common; sustained timeouts suggest network blocking, DNS filtering, or connectivity issues.