A transparent overview of what each class of 0bps tool measures, calculates or reads — and what it cannot prove.
Speed and Internet Quality tests use browser transfers against Cloudflare edge infrastructure. Multiple transfer sizes help avoid estimating a fast connection from one tiny payload. While download/upload transfers run, the same engine can also probe latency under load so 0bps can show the measured responsiveness cost of saturating the path. Results still depend on route, browser, device, VPN and local network conditions.
Speed, Internet Quality, Ping and Jitter use the same Cloudflare edge measurement engine for unloaded latency. Speed and Internet Quality additionally request simultaneous latency samples during active download/upload transfers. 0bps reports the positive increase above the idle baseline as an advisory load-latency / bufferbloat signal. These are HTTP edge measurements rather than native ICMP, so another destination can still differ.
Transfer, bitrate and data-usage calculators use deterministic arithmetic. Real traffic can differ because of protocol overhead, server limits, adaptive streaming, codec behavior and shared network load.
Audio and video checker modules read supported local-file metadata in the browser. Files are not intentionally uploaded to 0bps for these checks. Codec and container support varies by browser.
DNS Lookup and Reverse DNS use the server DNS resolver for forward records or PTR data. Resolver caches, TTLs and geography can produce answers that differ from another resolver or location. Reverse DNS performs a DNS query only and does not connect to the entered IP.
The 0bps Connection Score is a documented heuristic that combines throughput and responsiveness. It is intended to make raw results easier to read, not to certify an ISP, SLA or application guarantee.
Speed dial movement, quality-stage progress, Ping radar pulses and Jitter waveform points are driven by measurement-engine result changes or real latency samples already produced by the active test. The Speed dial starts at a standard 100 Mbps range, can expand from real download evidence, and keeps that range for upload so phase positions remain comparable. Its needle is contained inside the SVG geometry and visual inertia only interpolates between real callbacks; it does not create measurement samples. 0bps does not generate random throughput, ping or jitter values to make an animation look busy. Decorative motion is reduced when the browser reports a reduced-motion preference.
For troubleshooting, compare repeated measurements under similar conditions. Change one variable at a time — Wi-Fi vs Ethernet, VPN on vs off, router position, time of day or device — before drawing conclusions.
Browser-to-edge throughput, not a guarantee for every destination.
Not native ICMP ping. Real service routes can differ.
Advisory bufferbloat signal; not router queue telemetry.
Variation across the final bounded callback updates; visual interpolation is never counted as evidence.
Download 35%, upload 20%, unloaded latency 30%, jitter 15%. Loaded latency remains outside the score.
No account sync. Export occurs only when you explicitly request a local file.
v1.44 connects measured history with planning tools. USE LAST TEST RESULT only copies a stored browser-local measurement into an applicable calculator field; it never silently changes plan values or sends the measurement to the server.