Gamepad Tester — Online Controller Tester (Buttons, Drift, Vibration)
Plug in or pair your controller, then press any button — browsers only show a gamepad after its first input. This controller tester online lights up all 17 standard buttons, shows both sticks with raw four-decimal axis values and analog trigger bars, runs a 5-second joystick drift test, measures how far each stick actually reaches, and rumbles the motors when the browser allows it. It works with Xbox, PlayStation and Switch-style pads and generic USB controllers.
Private by design. Controller input is read from the browser's Gamepad API and processed in this tab only — nothing is uploaded, recorded or sent to a server, and the optional test report is copied to your own clipboard.
How to test a gamepad online
- Connect the controller and press a button. Plug it in by USB or pair it over Bluetooth, keep this tab in front, then press any button or move a stick. The browser reveals the pad on that first input and it appears in the controller list.
- Press every button and move both sticks. Each button highlights on the diagram as you press it and is marked as tested in the list. Check that triggers show a smooth 0 to 1 value instead of jumping, and that each stick dot follows your thumb.
- Run the 5-second drift test. Put the controller on a table, click Start drift test and do not touch it. The page records every resting reading, reports mean offset and maximum radius per stick, and gives a verdict from Normal (up to 5% of travel) to Severe.
- Check range and vibration, then copy the report. Run the range test and rotate each stick around its rim for 5 seconds to see how evenly it reaches every direction. Use the weak and strong rumble buttons to test the motors, then copy the text report if you need to share the results.
Use Cases
Frequently Asked Questions
Why is my controller not detected?
First press a button or move a stick with this tab in front: browsers do not list a gamepad until it sends input, so a connected but untouched pad stays invisible. Then check the connection. Wired pads need a data cable (some cables only charge). For Bluetooth, the pad must be paired in the operating system, not just in another app. Close Steam, Steam Input, a console-companion app, an emulator or any game that may have grabbed the controller, since an exclusive grab hides it from the browser. Chrome and Edge list most pads in the standard layout. Firefox may show some pads with a non-standard layout, which this page handles with a raw index table. Safari supports only controllers the operating system recognises. Finally, make sure the page is open over https in its own tab, not inside an embedded frame.
Does this work as a PS5, Xbox or Switch Pro controller tester?
It tests any pad the browser exposes, including DualSense, Xbox One and Series controllers and Switch Pro style pads, over USB or Bluetooth. The diagram labels buttons by index rather than by console symbol: with the standard layout, button 0 is the bottom face button and 3 is the top one, whatever its printed symbol. Extras such as a touchpad click, mic mute, share or capture buttons, gyro, adaptive triggers and speaker audio are not part of the standard layout, so they may appear as extra button indexes in some browsers or not at all. Pads that connect with a non-standard mapping are shown as raw buttons[] and axes[] tables.
What is stick drift, and what dead zone do games apply?
Drift is a stick reporting movement while nobody touches it, usually from worn contacts, dust or a sensor that no longer rests at centre. Games hide small resting noise with a dead zone, a circle around the centre where input is treated as zero; settings often fall between about 5% and 25% of full travel, and Microsoft's XInput headers suggest roughly 24% for the left stick and 27% for the right. This page grades the largest resting deflection: up to 5% is Normal, up to 10% Minor, up to 20% Noticeable, up to 30% Significant and anything above that Severe. It also tells you the smallest dead zone that would have hidden every reading.
How reliable is the joystick drift test?
It samples the stick on every screen frame for 5 seconds, which is typically 300 to 600 readings depending on your display refresh rate, and reports mean offset (where the stick rests), maximum radius (the worst reading) and jitter (noise around the resting point). If the controller is touched or tilted on a soft surface the numbers will be wrong, so test on a table with hands off, and run it twice. The range test splits the stick circle into 36 sectors, takes the farthest point in each and reports smallest reach divided by largest reach; it needs at least 80% of sectors visited. 90% or more counts as even, 75% to 90% slightly uneven. Square-gate sticks can read above 1.0 in the corners, which is normal.
Why does the vibration test do nothing?
Vibration through a browser is limited. Chrome and Edge expose rumble through the Gamepad API for supported pads, often only over USB, while Firefox and Safari commonly expose none. If the Vibration card says it is unavailable, the browser is not offering the motors, which does not mean they are broken. Some controllers only rumble when a game claims them, or when another app is not holding the device. When the command is accepted, the page reports it as sent: whether you feel the weak (high-frequency) or strong (low-frequency) motor is the real test.
Is the update rate the same as my controller's polling rate?
No. The update rate card counts how often the timestamp of the gamepad state changes between our reads, one read per screen frame. Browsers batch and throttle input, a 60 Hz display caps the reads at 60 per second, and some browsers only change the timestamp when a value changes, so you must keep a stick moving to get a figure. Use it to spot a pad that stutters or updates rarely, not to measure the hardware polling rate of a 125 Hz, 250 Hz or 1000 Hz controller.
Stick drift verdicts used by this gamepad tester
Radius is the distance of the stick from centre in axis units, where 0 is perfectly centred and 1 is full deflection. The verdict uses the largest radius seen while the stick was untouched, because that is what a game dead zone has to cover. The last column shows what to do about it.
| Max radius | Share of full travel | Verdict | What it means | What to do |
|---|---|---|---|---|
| up to 0.05 | up to 5% | Normal | Within the noise of a healthy stick. Any dead zone, even a tight 5% one, hides it. | Nothing to do. |
| 0.05 to 0.10 | 5% to 10% | Minor drift | Hidden by a typical 10% dead zone. Games or apps that use a 5% dead zone may creep slightly. | Blow dust out, retest, or raise the dead zone a little. |
| 0.10 to 0.20 | 10% to 20% | Noticeable drift | Beyond a small dead zone: cameras pan slowly or menus scroll by themselves in games that use less than 20%. | Clean around the stick base, recalibrate if the pad supports it, raise the dead zone to match. |
| 0.20 to 0.30 | 20% to 30% | Significant drift | Only a large dead zone (20% to 30%) hides it, and that costs fine aiming near the centre. | Plan a repair or replacement of the stick module; check warranty. |
| above 0.30 | above 30% | Severe drift | No reasonable dead zone fixes it. The stick is worn out, stuck off-centre or damaged. | Repair or replace the stick module or the controller. |
- Healthy sticks rest within 5% of centre; a dead zone of 10% hides everything up to Minor drift, and 20% covers Noticeable drift at the cost of fine control near the centre.
- Range test: the stick circle is cut into 36 sectors and circularity is the smallest sector reach divided by the largest. 90% or higher counts as even range; 75% to 90% is slightly uneven; below 75% suggests a damaged or obstructed stick.
- Both tests read the first four axes (axes 0 and 1 for the left stick, 2 and 3 for the right). On pads with a non-standard mapping, check the raw axes table to confirm which axes are the sticks.
Related tools
More free, in-browser tools from the Device Tests set — every one runs locally, with no sign-up and no upload.