What is the Steam Controller polling rate on PC?
250 Hz on the USB cable and on the Puck, about 130 Hz over Bluetooth. That is a quarter of the 1000 Hz pads on our other pages, but the measured latency is better than the rate suggests: the Steam Controller's wired buttons beat a stock DualSense at the same 250 Hz by about 4 ms, and the Puck costs only 2 to 3 ms over the cable.
The 2026 Steam Controller polls at 250 Hz on PC over its USB cable and over the Puck, and at about 130 Hz over Bluetooth. Gamepadla measured 241.49 to 256 Hz on the cable, 258.46 to 271 Hz on the Puck and 129 to 134.23 Hz on Bluetooth, and tests on firmware 0926 read exactly 250 Hz on both the cable and the Puck. Valve's spec says the same thing in different units. That is a quarter of the roughly 1000 Hz of the wired pads on our GameSir polling rate page. It is not a slow controller, though: on buttons it measured 4.12 ms wired and 6.25 ms on the Puck.
What "4ms polling rate" means
Valve's Steam Controller tech specs list, under the Steam Controller Puck: "2.4GHz Wireless connection ~8ms full end-to-end, 4ms polling rate (measured at 5m)". A polling rate given in milliseconds is a report interval, and 1000 / 4 = 250 Hz. Valve gives no rate for the cable or Bluetooth; the spec lists "USB-C tethered play" and "Bluetooth 4.2 minimum, 5.0 or higher recommended", nothing more.
So the measurement and the spec agree. On the half-interval rule from our polling rate page, 250 Hz costs 2.0 ms of mean polling delay against 0.5 ms at 1000 Hz, a difference of 1.5 ms (derived). Over Bluetooth at 134.23 Hz the interval is 7.45 ms and the mean delay 3.72 ms.
Every link, measured
| Link and mode | Polling (Hz) | Buttons (ms) | Sticks (ms) |
|---|---|---|---|
| Cable, Steam | 241.49 | 4.12 | 5.08 |
| Cable, XInput | 256 | 6.60 | 4.07 |
| Puck, Steam | 258.46 | 6.25 | 7.75 |
| Puck, XInput | 271 | 9.14 | 7.35 |
| Bluetooth, Steam | 134.23 | 7.26 | 10.21 |
| Bluetooth, XInput | 129 | 9.76 | 8.04 |
The polling tool's average rates were 248.95 Hz on the cable, 269.14 Hz on the Puck and 133.42 Hz on Bluetooth, and gamepadla's July stick tests logged a 248 Hz report rate on the cable in both Steam and XInput modes. Every reading sits close to 250 Hz, or about half that on Bluetooth.
The Puck costs 2 to 3 ms, by two methods
Puck minus cable, same tester and day (derived): 2.13 ms on buttons and 2.67 ms on sticks in Steam mode, 2.54 and 3.28 ms in XInput mode. A third tester on firmware 0926, in October 2026, measured sticks at 2.75 ms on the cable and 5.41 ms on the Puck, both at 250 Hz, a 2.66 ms gap.
Gamers Nexus reached the same answer with a different method. They soldered leads to one of the controller's inputs, drove it from an NVIDIA LDAT and timed button press to light on the screen. Over 499 clicks wired averaged 19 ms with a 3.1 ms standard deviation, and the Puck 21.6 ms with the same 3.1. That is a 2.6 ms gap (derived), inside the range gamepadla's testers found. A second Steam Controller on the same Puck left it at 21.9 ms, and a USB Xbox One controller (model 1537) scored 22.6 ms.
That also settles the two latency numbers in circulation. Valve's "~8ms full end-to-end" and Gamers Nexus's 21.6 ms are not in conflict: Gamers Nexus timed the whole system, where a ROG Keris mouse scored 15 ms on the same rig. Gamepadla times the pad and its link, and its Puck averages of 6.25 to 9.14 ms on buttons and 7.35 to 7.75 ms on sticks sit around Valve's figure.
Bluetooth is the one to avoid
On gamepadla's rig Bluetooth added 3.14 ms on buttons and 5.13 ms on sticks over the cable in Steam mode (derived). Gamers Nexus, with the pad next to the test system on a ROG Maximus Z890 Hero's built-in radio, measured 37.3 ms with a 20.6 ms standard deviation: 18.3 ms slower than wired (derived) and far less consistent. With seven other controllers on Bluetooth, the Steam Controller averaged 73.8 ms with 101 missing entries, while the Puck in the same crowd held 21.8 ms.
250 Hz, but not slow
A low polling rate does not make a slow pad. At the same 250 Hz, a stock wired DualSense measured 8.27 ms on buttons and 7.41 on sticks; the Steam Controller's cable in Steam mode is 4.15 and 2.33 ms faster (derived). A Microsoft pad at 125 Hz, on our Xbox controller polling rate page, measured 6.50 and 8.30 ms wired. Against 1000 Hz pads it gives ground on buttons and not on sticks: the 8BitDo Ultimate 2 Wireless on our 8BitDo Ultimate polling rate page measured 2.81 ms on buttons but 7.02 ms on sticks, 1.94 ms slower on sticks than the Steam Controller.
The Switch 2 Pro Controller, the other first-party pad near 250 Hz on PC, measured 7.37 to 7.97 ms on buttons on a cable outside Switch mode, on our Switch Pro Controller polling rate page. The Steam Controller's 4.12 ms is 3.25 to 3.85 ms quicker (derived).
Gyro at 250 Hz
Gyro is the complaint that follows the 250 Hz figure. Valve publishes no separate gyro rate, only a "6-axis IMU". At 250 Hz a report arrives every 4 ms, so (derived):
- 144 fps: 1.74 reports per frame, every frame gets new data.
- 240 fps: 1.04 reports per frame, about one each.
- 360 fps: 0.69 reports per frame, so about one frame in three gets no new movement.
- Bluetooth at 134 Hz and 144 fps: 0.93, so some frames already repeat.
Valve has also acknowledged a real stutter. A May 2026 Steam beta, as published by GamingOnLinux, lists "Steam Controller: Minor improvement to gyro micro stutter (ignore clone packets)". If gyro feels choppy, update Steam first, use the cable or Puck rather than Bluetooth, and check whether your frame rate is above 250.
Where this stops applying
All latency figures are Windows. Each gamepadla row is one unit tested by one person, and differences under 1 ms sit inside its stated margin. The two testers used different Windows versions and modes on different days, so compare within a tester. The October firmware 0926 stick results, from a different tester, came out about half the earlier ones (2.75 against 5.08 ms on the cable in Steam mode); gamepadla gives no reason, and a new firmware or Steam update can move any of these numbers. Gamers Nexus's figures include the game, GPU and display, so only the differences between their links compare with gamepadla's. None of this covers stick precision, which comes down to the sensor and your deadzone.
What to do
- Competitive play on PC: the cable. 250 Hz, 4.12 ms on buttons in Steam mode.
- Wireless: the Puck. It added 2 to 3 ms over the cable by both gamepadla and Gamers Nexus.
- Avoid Bluetooth: about 130 Hz, and 37.3 ms click to photon with a 20.6 ms spread.
- Gyro aiming above 250 fps: expect frames without a fresh report; keep Steam updated.
- Need 1000 Hz on paper: buy a different pad. Valve lists no faster rate.
Frequently asked questions
Is the Steam Controller 1000 Hz?
No. Gamepadla measured 241.49 to 256 Hz on the cable, 258.46 to 271 Hz on the Puck and 129 to 134.23 Hz over Bluetooth on Windows. Newer tests on firmware 0926 in October 2026 read exactly 250 Hz on both the cable and the Puck. Valve's own spec gives the Puck a "4ms polling rate", which is a 4 ms report interval, or 250 Hz.
Is the Steam Controller Puck as fast as a cable?
Almost. In gamepadla's tests the Puck added 2.13 ms on buttons and 2.67 ms on sticks over the cable for one tester, and 2.54 and 3.28 ms for another. Gamers Nexus measured 21.6 ms click to photon on the Puck against 19 ms wired, a 2.6 ms gap with the same 3.1 ms standard deviation.
Why does Valve say 8 ms when Gamers Nexus measured 21.6 ms?
They measure different spans. Gamers Nexus timed the whole system, from button press to light on the screen, and a high-end gaming mouse scored 15 ms on the same rig. Gamepadla's Puck averages, which time the controller and its link only, were 6.25 to 9.14 ms on buttons and 7.35 to 7.75 ms on sticks, around Valve's ~8 ms.
Is Bluetooth bad on the Steam Controller?
It is the slowest link. It polls at about 130 Hz, and Gamers Nexus measured 37.3 ms click to photon with a 20.6 ms standard deviation, against 19 ms wired. With seven other Bluetooth controllers nearby it averaged 73.8 ms with 101 missing entries. Use the Puck instead: it held 21.8 ms in the same crowded test.
Does 250 Hz make Steam Controller gyro choppy?
Valve has acknowledged gyro stutter: a May 2026 Steam beta listed a "Minor improvement to gyro micro stutter (ignore clone packets)". Valve publishes no separate gyro rate. At 250 Hz the pad sends a report every 4 ms, about one per frame at 240 fps, and at 360 fps roughly one frame in three gets no new report.