Cable-heavy footprint with some fiber
Xfinity speed test
Separate device and Wi-Fi limits from the Xfinity connection by repeating one controlled browser test on Ethernet and Wi-Fi.
Appears in 21 of 50 FCC-derived state profiles
Independent ISP diagnostics
Internet speed tests by provider
Run one transparent test engine with the provider, access technology, official diagnostic, and controlled comparison steps already in view. These are useful provider workflows, not provider-owned tests or generic keyword pages.
10 provider workflowsProvider context savedFCC-derived state entry pointsOpen JSON + CSV
Independent scope
Every result includes the browser, device, local network, ISP path, internet route, and selected SwiftSpeedTest endpoint. It does not isolate the provider access line or read gateway telemetry.
No affiliation claim
Provider names and trademarks belong to their owners. SwiftSpeedTest is independent and links to provider-operated diagnostics so readers can compare measurement boundaries.
Choose the connection you use
Provider-specific test workflows
Each page pre-fills the provider in saved results and changes the interpretation workflow for cable, fiber, copper, fixed wireless, or satellite service.
Cable-heavy footprint with some fiber
Spectrum speed test
Test a Spectrum connection with a controlled wired baseline, a room-level Wi-Fi repeat, and saved time-of-day evidence.
Appears in 14 of 50 FCC-derived state profiles
Mixed fiber, copper, and fixed-wireless footprint
AT&T speed test
Record the exact AT&T access type before interpreting a browser result because fiber, copper, and fixed wireless have different bottlenecks.
Appears in 27 of 50 FCC-derived state profiles
Fiber and fixed-wireless footprint
Verizon speed test
Interpret Verizon results by access type: Fios, 5G Home, and LTE Home should be tested and compared separately.
Appears in 43 of 50 FCC-derived state profiles
Fixed-wireless footprint
T-Mobile Home Internet speed test
Track T-Mobile Home Internet with stable gateway placement, signal-aware repeats, and busy-versus-quiet time windows.
Appears in 49 of 50 FCC-derived state profiles
Cable-heavy footprint with some fiber
Cox speed test
Build a Cox evidence set with wired and Wi-Fi comparisons, upload checks, loaded latency, and repeatable time windows.
Appears in 8 of 50 FCC-derived state profiles
Mixed copper and fiber footprint
CenturyLink speed test
Record whether the address uses CenturyLink copper or fiber before judging a device result or comparing it with a plan.
Appears in 15 of 50 FCC-derived state profiles
Mixed fiber and copper footprint
Frontier speed test
Separate Frontier Fiber or copper access from the device and Wi-Fi path, then keep comparable runs in a private history.
Appears in 2 of 50 FCC-derived state profiles
Cable and fiber footprint
Optimum speed test
Identify Optimum cable or fiber service, then compare wired, Wi-Fi, loaded-latency, and time-of-day evidence.
Appears in 2 of 50 FCC-derived state profiles
Low-Earth-orbit satellite option
Starlink speed test
Track Starlink with stable dish and router conditions, obstruction-aware app checks, and repeated latency evidence.
Appears in 50 of 50 FCC-derived state profiles
Versioned method 1.0
How the provider context is built
1. One consistent measurement engine
Provider pages use the same adaptive browser transfer, idle and loaded latency, jitter, endpoint, and request-confidence method.
2. Access-specific interpretation
The workflow changes for cable, fiber, copper, fixed wireless, and satellite without publishing fabricated plan targets or provider-wide rankings.
3. FCC-derived state context
State links come from provider-reported fixed broadband availability as of December 31, 2025. They are not market share or address eligibility.
4. Reusable, citable releases
The profile definitions, official resources, limitations, and state presence are published as current and immutable JSON plus CSV.
Frequently asked questions
Questions about independent provider speed tests
- Are these speed tests operated by the internet providers?
- No. SwiftSpeedTest is independent and is not sponsored by, endorsed by, or affiliated with the listed providers. Each page runs the same transparent browser measurement and links to provider-owned tools when a different gateway, router, app, or account-level view is useful.
- What changes between the provider test pages?
- The measurement engine stays consistent. The saved provider context, access-technology cautions, controlled checks, official diagnostic links, and FCC-derived state entry points change so the workflow fits the connection being investigated.
- Does a provider page measure the ISP access line directly?
- No. The browser result combines the test device, browser, local network, ISP path, internet route, and selected SwiftSpeedTest endpoint. It cannot isolate speed reaching a gateway or reproduce a provider-owned network diagnostic.
- Why publish a versioned provider profile dataset?
- The JSON and CSV releases expose the measurement boundary, diagnostic factors, controlled steps, official resources, and FCC-derived state presence in a form researchers and answer engines can inspect, cite, and compare over time.
- Do the FCC state counts show market share or address availability?
- No. They only show where a provider appears in SwiftSpeedTest's state profiles built from provider-reported FCC fixed broadband availability. Footprints can overlap, the percentages are not market share, and service must be checked at the exact address.
- How many results should I collect before contacting support?
- One test is a snapshot. Save several full-confidence runs from the same device and endpoint, including wired and Wi-Fi checks plus quiet and busy time windows. The history and plan-audit tools can then summarize medians, spread, and test conditions without claiming a cause.
Trademark notice: AT&T, CenturyLink, Cox, Frontier, Optimum, Spectrum, Starlink, T-Mobile, Verizon, Xfinity, and related marks are property of their respective owners. Listing a provider describes user-entered test context and public source material; it does not imply affiliation.