A que velocidade vou?

1. What this speedometer does — and what it deliberately does not do

This page is a live measuring instrument, not a blog post with a screenshot. Tap Start, allow location once, and your browser reports your movement in real time: a large current-speed readout, average and maximum speed for the session, total distance, elapsed time, and a GPS accuracy badge in meters. Everything runs in the page itself — there is no app to install, no account to create, and no trip history uploaded anywhere. When you close or reset the tab, the session is gone. That design is intentional. Most speedometer apps ask for an account, background location, and cloud sync for a job that can be done locally in a few kilobytes of JavaScript. We chose the opposite: one page, one job, honest signal reporting.

What you will not find here matters just as much. We never invent a number when the signal is gone — in a tunnel, underground, or deep indoors the page shows Searching or Weak instead of a fake steady figure. We do not smooth away reality either: the raw fixes are averaged over a short window (typically the last three to four positions) so a single reflected signal does not send the readout jumping from 40 to 90 km/h, but sustained movement always shows through within a second or two. And we never present GPS as a certified instrument. It is an excellent reference — often closer to true speed than a car dashboard at steady cruise — but it remains a consumer sensor with known limits, which sections 3, 4 and 10 spell out plainly.

2. How your phone actually measures speed

Your phone does not have a speed sensor. It has a satellite receiver. At least four GPS satellites (often plus Galileo, GLONASS or BeiDou satellites — your phone uses whichever constellation is strongest) broadcast precise time stamps from atomic clocks. By comparing when each signal arrives, the phone solves its position in three dimensions. That calculation is called trilateration, and it repeats roughly once per second while you are tracking. Speed is then derived from those positions in one of two ways.

• Direct Doppler speed (preferred). Modern phone chips measure the Doppler shift of the satellite carrier wave — the same physics that makes an ambulance siren change pitch as it passes. Because this measures velocity directly from the radio signal, it can be remarkably stable even when your mapped dot wanders a few meters. When your browser exposes this value as coords.speed, this page shows it first.
• Position-difference fallback (Haversine). Some browsers and older devices report position without a speed field. In that case the page computes the great-circle distance between successive fixes with the Haversine formula and divides by elapsed time. A short moving average tames single-fix jumps. This fallback is slightly noisier at walking pace but converges well once you move steadily above about 5 km/h.

Here is the counterintuitive part worth remembering: speed can be more accurate than position. Your blue dot may sit five meters off the road on the map, yet your speed can still be right within 1–2 km/h. Position needs absolute accuracy; speed only needs thechange between fixes to be roughly right, and errors that shift both fixes the same way largely cancel out. That is why experienced cyclists trust a GPS average over a ten-second stretch even when the live dot looks jittery — the section on best practice below builds on exactly this insight.

3. How accurate is phone GPS speed, really?

Short answer: outdoors with open sky, at steady speed, expect within about 1–3 km/h of true speed on a modern phone. That is a typical field observation, not a laboratory guarantee — your hardware, browser, satellite geometry and surroundings all matter. The page helps you judge each moment with the accuracy badge: ± meters is the browser's own estimate of position uncertainty, and it correlates strongly with speed trust. Use this table as a rule of thumb:

Two practical consequences follow. First, judge averages, not spikes. A one-second jump to 68 km/h while cycling at 25 km/h is almost certainly a reflected signal, not a sprint — the session average over ten seconds or more is the honest number. Second,wait for the badge to settle. The first 20–30 seconds after tapping Start are warm-up: the receiver is acquiring satellites and the accuracy figure improves from ±40 m toward ±8 m. Readings taken during warm-up deserve skepticism; readings after it deserve confidence.

4. GPS vs your car dashboard vs a bike computer vs a phone app

People land here asking whether GPS is “better” than the speedometer they already have. The honest answer depends on which one you mean:

The car-dashboard detail surprises most drivers: in the EU, US and India, regulations allow — and manufacturers use — a deliberate optimistic bias, so 100 km/h on the dial is often 94–98 km/h of true speed. That is a safety margin, not a fault. At a steady 90 km/h on open highway with a green accuracy badge, this GPS page is frequently closer to reality than the dial. But the dial wins everywhere GPS cannot see the sky, and it remains the only legal instrument. Use GPS as a respected second opinion: if both agree within a few km/h at cruise, both are healthy; if they disagree by more than 10% repeatedly in good signal, suspect tire size, a failing wheel sensor, or a phone case blocking the antenna — and get the car checked rather than assuming the phone is wrong.

5. Units explained: km/h, mph, m/s and knots

Speed is speed, but the number you want depends on where you are and what you are doing. This page offers four units and remembers your choice on your device only. Kilometers per hour (km/h) is the road standard across most of the world; miles per hour (mph) is used on roads in the US, UK and a few other countries; meters per second (m/s) is the scientific SI unit beloved by runners, physicists and wind readings; knots (nautical miles per hour) is the marine and aviation standard. The boat page on this site defaults to knots for exactly that reason — sailors think in knots the way drivers think in km/h.

Conversions are exact and use the international definitions: 1 mph = 1.609344 km/h, 1 m/s = 3.6 km/h, and 1 knot = 1.852 km/h. So 50 km/h is 31.1 mph, 13.9 m/s, or 27.0 knots; 60 mph is 96.6 km/h; 100 km/h is 62.1 mph or 54.0 knots. The reference table above the guide lists common equivalents, and the dedicated MPH → KPH and KPH → MPH converters do live math for any value you type. A common mistake is converting average and max separately with rounded intermediates — convert from the original figure each time, or let the unit switcher do it, which converts from the underlying meters-per-second value without rounding drift.

6. How to get the most accurate reading: eight habits that actually help

These are ordered by impact — the first three fix the majority of “inaccurate GPS” complaints:

1. 1. Go outdoors with open sky. Buildings, roofs, and even a car windshield coating are the biggest error sources. A footpath beats a balcony; a park beats a downtown canyon. If you must be inside, sit by a window and accept indicative numbers.
2. 2. Hold steady or mount — never swing. A phone swinging in your hand adds meters of phantom movement per second. Pocket, handbag or swinging arm all inject false speed. Mount on handlebars or dashboard, or hold flat and still at chest height.
3. 3. Wait 20–30 seconds after Start. Watch the ±m badge fall from ±30 m toward ±10 m or better before trusting the big number. The session average and max are most meaningful after warm-up — reset once the badge is green if you started cold.
4. 4. Keep high-accuracy mode on. On Android, set Location to High accuracy (GPS + Wi-Fi + mobile networks). On iPhone, keep Location Services on and grant this site Precise Location. Power-saving location modes that use Wi-Fi only will degrade every reading.
5. 5. Judge the average, not the spike. At walking pace the fallback computation is inherently jumpy — a single reflected fix can briefly read 9 km/h while you stroll at 5. The average over 10+ seconds is the truth; the spike is noise. Same at 100 km/h: cruise average beats any one-second peak.
6. 6. Mind heat and battery savers. A hot phone throttles GPS updates, and aggressive battery savers kill background fixes. Lower screen brightness, close heavy apps, and exempt your browser from battery optimization during a long ride if readings start freezing.
7. 7. Remove signal blockers. Thick metal phone cases, a phone buried under luggage on a train, or a car with an athermic (heat-reflective) windshield can each cost 10–20 m of accuracy. Window seat, phone near glass, case off for the test — small changes, large effect.
8. 8. Calibrate against a known steady speed. Once, on open road as a passenger, compare ten seconds of GPS average against the car at a locked cruise speed. Note the offset (e.g. dial 100, GPS 96). That single experiment teaches you more about your devices than any forum thread.

7. Practical use cases: getting value from each mode

8. Privacy, permissions and battery: what actually happens

Permission prompts make people nervous, reasonably so. Here is exactly what happens on this site. When you tap Start, your browser — not us — asks whether to share location with this page. If you allow, coordinates flow from your phone's GPS chip to JavaScript running in this tab, where current, average, max, distance and accuracy are computed in memory. Those numbers are never sent to our servers, never written to a database, never handed to advertisers, and never included in analytics. Deny the prompt and nothing is collected at all; the page simply shows how to re-enable later. Close the tab and the session evaporates. Preferences you explicitly set — unit, theme, keep-awake — live in your browser's local storage on your device only. That is the entire data story, and the privacy policystates it in full legal detail.

Battery is the honest cost. Continuous satellite fixes plus an awake screen is among the hungriest things a phone does — expect roughly 10–20% drain per hour depending on device, brightness and heat. Three habits keep it manageable: keep Keep screen awake off unless you are actively glancing at the readout; dim the screen outdoors rather than maxing it; and pause tracking on long stops instead of letting elapsed time and GPS run idle. If the phone gets hot, give it shade — thermal throttling slows fixes before any setting does. And if your browser offers “precise” versus “approximate” location, choose precise for this tool; approximate location (often Wi-Fi-based, ±500 m or worse) cannot produce meaningful speed at all.

9. Troubleshooting: when the number looks wrong

Almost every report we receive falls into one of these buckets. Work through them in order:

Still stuck after these steps? Email hello@gpsspeedometer.bondwith your device, browser, whether you were outdoors or in a vehicle, and what the accuracy badge showed — that context lets us reproduce the issue instead of guessing. The contact page lists exactly what to include.

10. Limitations, safety and how this guide stays honest

A trustworthy instrument states its limits plainly, so here are ours. This site is a helpful reference, not a certified, calibrated or safety-critical device. It is not a replacement for your vehicle's legally required speedometer, a marine chartplotter, or any aviation instrument. Never hold or operate your phone while driving, riding or boating: mount it before moving, start tracking before departure or ask a passenger, and obey traffic and navigation law. GPS degrades in tunnels, underground, dense urban canyons, heavy tree cover and indoors — in those places the page shows weak or searching states rather than a confident fiction. Speed over ground is what GPS reports; it knows nothing of speed limits, currents, tire size or calibration certificates.

How we keep this page honest: the prose above describes the shipped code, not a marketing ideal — Doppler-first with Haversine fallback, short-window smoothing, green/amber/searching thresholds tied to the browser's own accuracy estimate. When browsers change Geolocation behavior or we adjust thresholds, this guide and thetechnical explainer are updated with a new review date. If you spot an error, a stale browser step, or a reading that contradicts this guide in good signal, tell us — corrections from real field reports are how a reference tool earns trust. Thank you for reading this far; now scroll up, tap Start outdoors, and watch an honest number move.