What is the GameSir controller polling rate on PC?
About 1000 Hz on the cable for 14 of the 16 GameSir controllers measured on PC, and close to 8000 Hz for the two 8K models. The trap is the dongle. It usually reports at the same rate, so the spec sheet looks identical, but gamepadla measured button latency two to four times higher on it than on the cable.
Most GameSir controllers poll at about 1000 Hz on the cable, and the two 8K models at close to 8000 Hz: 14 of the 16 in our table, with the older T4 Cyclone (500 Hz) and T4 Pro (149 Hz) the exceptions. Gamepadla measured the Cyclone 2 at 1000 Hz wired and 999 Hz on its 2.4G dongle under Windows 11, which looks like the same result twice. It is not: average button latency was 2.40 ms on the cable and 7.40 ms on the dongle. With GameSir, the polling rate on the box tells you very little about which link to use.
Sixteen GameSir pads, measured
The table gives one primary test per link, Windows, XInput mode unless noted. The stated column quotes the controller's page on gamesir.com where that page gives a figure.
| Controller and link | GameSir states | Polling (Hz) | Buttons (ms) | Sticks (ms) |
|---|---|---|---|---|
| Tarantula 8K, cable | 8000 Hz wired | 8000 | 1.04 | 0.49 |
| G7 Pro 8K Aimlabs Edition, cable | 8000 Hz wired and 2.4G | 7936.51 | 1.39 | 0.69 |
| G7 Pro 8K, dongle | 8000 Hz wired and 2.4G | 7936.51 | 3.47 | 3.56 |
| G7 Pro 8K, Bluetooth (Sony mode) | - | 800 | 3.58 | 3.45 |
| Stargazer, cable | - | 999 | 2.14 | 2.34 |
| Stargazer, dongle | - | 998 | 5.70 | 5.97 |
| Stargazer, Bluetooth (DInput) | - | 100.55 | 15.1 | 15.7 |
| Cyclone 2, cable | - | 1000 | 2.40 | 6.51 |
| Cyclone 2, dongle | - | 999 | 7.40 | 17.0 |
| Cyclone 2, Bluetooth (Switch mode) | - | 399.2 | 6.01 | 10.3 |
| G7 Pro, cable | 1000 Hz wired and 2.4G | 1001 | 2.43 | 2.70 |
| G7 Pro, dongle | 1000 Hz wired and 2.4G | 1000 | 5.69 | 15.0 |
| G7 Pro, Bluetooth (DInput) | - | 400.32 | 5.37 | 5.82 |
| Nova 2 Lite, cable | - | 959.69 | 2.36 | 7.99 |
| Nova 2 Lite, dongle | - | 928.51 | 12.7 | 17.1 |
| Nova 2 Lite, Bluetooth (Switch mode) | - | 198.53 | 12.3 | 17.0 |
| T4 Kaleid, cable | - | 1001 | 2.61 | 8.49 |
| Tegenaria Lite, cable | - | 999 | 2.82 | 1.92 |
| G7 HE, cable | - | 987.92 | 2.84 | 11.6 |
| Tarantula Pro for Xbox, cable | 1000 Hz on PC | 1005.95 | 2.97 | 2.33 |
| G7 SE, cable | 1000 Hz on PC | 1000 | 3.00 | 3.23 |
| K1 Kaleid for Xbox, cable | - | 1000 | 5.32 | 7.62 |
| Tarantula Pro (tri-mode), cable | - | 1000 | 7.83 | 12.7 |
| Tarantula Pro (tri-mode), dongle | - | 1000 | 31.7 | 29.9 |
| Tarantula Pro (tri-mode), Bluetooth (Sony mode) | - | 162.13 | 19.2 | 25.4 |
| T4 Cyclone, cable | - | 499.75 | 6.80 | 46.1 |
| T4 Cyclone, dongle | - | 250.19 | 12.7 | 68.6 |
| T4 Pro, cable | - | 149.17 | 8.58 | 12.0 |
| Super Nova, cable (GPDL) | - | 996.02 | 7.32 | - |
| Super Nova, dongle (GPDL) | - | 999 | 20.4 | - |
Test details: Tarantula 8K, firmware 2.2.9, 14 July 2026. G7 Pro 8K, firmware 2.1.0, 17 and 18 May 2026. Stargazer, 19 August 2026. Cyclone 2, firmware 3.5.2, 21 April 2026. G7 Pro, firmware 2.3.9, 22 April 2026. Nova 2 Lite, 13 December 2025. T4 Kaleid, 7 June 2026. Tegenaria Lite, 26 April 2026. G7 HE, 20 March 2025. Tarantula Pro for Xbox, 24 July 2026. G7 SE, firmware 6.6.4, 10 June 2026. K1 Kaleid, firmware 1.49, 30 September 2024. Tarantula Pro (tri-mode), 4 May 2026. T4 Cyclone, 17 June 2026. T4 Pro, 3 April 2025. Super Nova, 1 March 2025. Gamepadla lists the tri-mode Tarantula Pro as "GameSir Tarantula Pro" on a separate page from the wired Xbox model. Stated figures are from GameSir's pages for the Tarantula 8K PC, G7 Pro 8K Aimlabs Edition, G7 Pro, Tarantula Pro for Xbox and G7 SE; the Cyclone 2 page gives no polling figure in its text.
Same rate on the dongle, not the same latency
Seven pads were tested over both links. On six, the dongle reported within 4% of the cable's rate; the T4 Cyclone's fell from 499.75 to 250.19 Hz. All seven were slower on buttons over the dongle. The difference is far larger than polling can explain: at 1000 Hz the mean wait for the next report is 0.5 ms on either link (derived, from the half-interval rule on our polling rate page). The rest sits in the wireless path.
| Controller | Cable (ms) | Dongle (ms) | Dongle costs (ms) |
|---|---|---|---|
| G7 Pro 8K | 1.39 | 3.47 | 2.08 |
| G7 Pro | 2.43 | 5.69 | 3.26 |
| Stargazer | 2.14 | 5.70 | 3.56 |
| Cyclone 2 | 2.40 | 7.40 | 5.00 |
| Nova 2 Lite | 2.36 | 12.7 | 10.34 |
| T4 Cyclone | 6.80 | 12.7 | 5.90 |
| Tarantula Pro (tri-mode) | 7.83 | 31.7 | 23.87 |
Sticks can be worse. The G7 Pro measured 2.70 ms on the cable and 15.0 ms on the dongle, both at about 1000 Hz. A second tester on Windows 10 got 7.29 ms on the Cyclone 2 dongle, so the gap is not one bad unit. That is why "1000 Hz wired and 2.4G" on a spec sheet does not settle the choice: it promises the rate, and the rate was delivered on every pad that makes that claim.
Lowering the polling setting costs the sticks most
Three pads were tested on the same cable, the same firmware and the same day at about 1000, 500 and 250 Hz. The half-interval rule predicts a 1.5 ms rise from 1000 to 250 Hz. Buttons roughly followed it. Sticks did not:
| Controller | ~1000 Hz (ms) | ~500 Hz (ms) | ~250 Hz (ms) |
|---|---|---|---|
| G7 HE | 11.6 | 18.0 | 31.8 |
| T4 Kaleid | 8.49 | 15.9 | 30.8 |
| K1 Kaleid | 7.62 | 14.2 | 27.2 |
Sticks gained 19.6 to 22.3 ms between the top and bottom setting, more than ten times the prediction (derived). Buttons on the same tests rose 1.64 ms on the G7 HE and 1.58 ms on the T4 Kaleid, and did not rise at all on the K1 (5.32 ms at 1000 Hz, 4.98 ms at 250). Gamepadla does not say why the sticks react this way, so we will not guess. The practical rule is simple: never lower the polling setting on a GameSir pad you aim with.
Is 8000 Hz worth it over 1000 Hz?
On the Tarantula 8K, measured on one cable at both settings, 8000 Hz averaged 1.04 ms on buttons and 0.49 ms on sticks; 1000 Hz averaged 1.85 ms and 1.36 ms. That saves 0.81 and 0.87 ms, about twice the 0.44 ms the half-interval rule predicts (derived). Real, but smaller than the cable-versus-dongle gap on any pad above. A wired 1000 Hz Stargazer at 2.14 ms is still about a third of a wired Xbox Series controller at 6.50 ms.
Switch mode can cut the rate
Mode behaviour differs by model. The Cyclone 2 held 1000 Hz and 2.32 ms on the cable in Switch mode, essentially the same as XInput. The G7 Pro 8K fell from 7936.51 Hz to 1000 Hz in Switch mode, still 1.96 ms. The Tegenaria Lite dropped to 250 Hz and 6.91 ms, against 999 Hz and 2.82 ms in XInput, and the T4 Kaleid to 124.98 Hz and 11.1 ms. The same penalty shows up on the 8BitDo Ultimate range and on Nintendo's own pads in our Switch Pro Controller polling rate guide. On PC, keep a GameSir pad in XInput.
Where this stops applying
Each row is one unit, tested by one person on one firmware, and gamepadla gives its rig a margin of plus or minus 1 ms, so gaps under a millisecond sit inside it. Pads were tested by different people, and the T4 Pro, G7 HE, K1 Kaleid and Super Nova results are from 2024 and 2025, so firmware may have moved since. Everything is Windows on PC; GameSir's Tarantula Pro for Xbox page states its 1000 Hz "on PC", and a PC test says nothing about a console. Dongle results depend on placement, and a USB 3.0 port next to the receiver can raise the noise floor, as our USB 3.0 interference page shows. Polling also says nothing about stick precision, which is set by the sensor and your deadzone.
What to do
- Use the cable. On every GameSir pad tested both ways it beat the dongle, by 2.08 ms (G7 Pro 8K) to 23.87 ms (tri-mode Tarantula Pro).
- If you must go wireless, pick a pad with a small measured dongle gap: G7 Pro 8K, G7 Pro or Stargazer, all under 6 ms on buttons.
- Leave the polling setting at its maximum. At 250 Hz, sticks measured about 20 ms slower on three models.
- Stay in XInput. Switch mode cut the Tegenaria Lite to 250 Hz and the T4 Kaleid to 125 Hz.
- 8000 Hz saves under a millisecond over 1000 Hz. Buy it for the low latency the 8K pads measured overall, not for the number.
Frequently asked questions
Is the GameSir Cyclone 2 1000 Hz?
Yes, on the cable and the dongle. Gamepadla measured 1000 Hz wired and 999 Hz on the dongle under Windows 11, but average button latency was 2.40 ms on the cable and 7.40 ms on the dongle. Over Bluetooth in Switch mode it measured 399.2 Hz and 6.01 ms.
Does the GameSir G7 Pro 8K really poll at 8000 Hz?
Gamepadla measured the G7 Pro 8K Aimlabs Edition at 7936.51 Hz on both the cable and the 2.4G dongle, which matches GameSir's claim of 8000 Hz "over both wired and 2.4G wireless connections". Buttons averaged 1.39 ms wired and 3.47 ms on the dongle.
Is the GameSir dongle as fast as the cable?
No. On all seven GameSir pads gamepadla tested on both links, the dongle was slower on buttons, even where it matched the cable's polling rate: 3.26 ms slower on the G7 Pro, 5.00 ms on the Cyclone 2 and 23.87 ms on the tri-mode Tarantula Pro. Use the cable when you can.
Should I lower the polling rate on a GameSir controller?
Not to save anything you can feel. On the G7 HE, dropping from about 1000 Hz to 250 Hz raised average stick latency from 11.6 ms to 31.8 ms in gamepadla's tests. The T4 Kaleid and K1 Kaleid showed the same pattern. Leave it at the highest setting.
What is the GameSir polling rate over Bluetooth?
It varies by model, from 100.55 Hz on the Stargazer to 800 Hz on the G7 Pro 8K. The Cyclone 2 measured 399.2 Hz and the G7 Pro 400.32 Hz. All of them were slower on buttons than their own cable.