What the score uses
Download, upload, unloaded latency and jitter remain the four score inputs. Loaded download/upload latency is reported separately so bufferbloat evidence is visible without silently changing the historical score model.
Run an edge-network measurement and translate throughput, latency and jitter into practical use-case ratings.
—
—
The score remains comparable with earlier 0bps results: download 35%, upload 20%, unloaded latency 30% and jitter 15%. Loaded latency is shown separately as diagnostic evidence and does not alter the score.
One score is one sample. Retest at different times; consistency across runs is more informative than a single isolated peak.
Loaded latency is additional measured evidence. The current Connection Score intentionally stays on the established four-signal model so v1.33 scores remain comparable with prior browser history. Use the load signal when diagnosing lag that appears only while the connection is busy.
Scores are deterministic 0bps heuristics intended to make raw measurements easier to interpret. They are not guarantees of application performance.
Download, upload, unloaded latency and jitter remain the four score inputs. Loaded download/upload latency is reported separately so bufferbloat evidence is visible without silently changing the historical score model.
Gaming cares more about latency and jitter; 4K streaming cares mostly about sustained download speed; live streaming depends strongly on upload capacity.
0bps never hides the raw measurements behind the score. The rating stays on the established download/upload/unloaded-latency/jitter model while latency-under-load is shown separately as additional diagnostic evidence.
The overall score summarizes one measured sample. Gaming, 4K streaming, calls and live streaming are rated separately because each workload weights the same measurements differently.
A connection can be excellent for streaming but only fair for gaming when throughput is high and responsiveness is weak. Use the workload closest to what you actually do.
The score is an explainable 0bps heuristic, not an ISP SLA, regulator metric or guarantee that a specific application will perform identically. Loaded-latency evidence is not silently folded into the score.
Retest when the network is busy and quiet. Large score swings usually matter more than a single isolated score because they reveal changing conditions.
0BPS SHOWS MEASURED OR CALCULATED FACTS SEPARATELY FROM INTERPRETATION. WHERE A RESULT IS ADVISORY, ESTIMATED OR DESTINATION-SPECIFIC, THE LIMIT IS STATED IN THE TOOL.
It is a transparent 0–100 interpretation of download speed, upload speed, latency and jitter. It is an advisory score, not an industry certification.
Gaming is more sensitive to latency and jitter, while 4K streaming depends more heavily on sustained download throughput. Live streaming also gives more weight to upload capacity.
No. Loaded latency remains separate measured diagnostic evidence so scores stay comparable with earlier browser history. The score still uses download, upload, unloaded latency and jitter.
The 0bps score combines download, upload, latency and jitter into a transparent planning heuristic.
Gaming favors latency and jitter; live streaming places more weight on upload capacity; 4K playback favors download capacity.
Short examples using the same units and assumptions as this module.
A moderate download rate can still score well for gaming when latency and jitter are low.
A fast download rate does not compensate for unstable latency in every real-time application.