What is the GuliKit KK3 Max polling rate on PC?
About 1000 Hz on the cable and 850 to 1000 Hz on the 2.4G dongle, so the spec sheet is right about the rate. It is wrong about what the rate buys you. On gamepadla's hardware rig the KK3 Max dongle was about 30 ms slower on buttons than the cable, in all four like-for-like sessions, by three testers on three firmware versions. GuliKit's newer Hyperlink 2 pads are the opposite case: plain Bluetooth at close to 800 Hz and under 5 ms.
The GuliKit KK3 Max polls at about 1000 Hz on PC over its cable and only a little less over its 2.4G dongle. Gamepadla measured 992.06 Hz wired under Windows 11, and 884.46 Hz on the dongle. So the 1000 Hz claim is close to true on both links, and it does not mean what buyers assume: average button latency was 5.40 ms on the cable and 35.4 ms on the dongle. The dongle is about 30 ms slower at nearly the same polling rate.
Eleven GuliKit pads, measured
GuliKit sells two generations that behave nothing alike. The KK3 family and the Elves 2 Pro use a 2.4G dongle or ordinary Bluetooth. The ES, ES Pro, Elves 2 and TT models use what GuliKit calls Hyperlink 2, a Bluetooth system built into the pad, and the ES Max ships with the Hyperlink 2 adapter, according to Notebookcheck. The table gives one primary test per link, XInput mode.
| Controller and link | Polling (Hz) | Buttons (ms) | Sticks (ms) |
|---|---|---|---|
| KK3 Max, cable | 992.06 | 5.40 | 33.9 |
| KK3 Max, 2.4G dongle | 884.46 | 35.4 | 53.1 |
| KK3 Max, Hyperlink 2 adapter | 249.19 | 18.3 | 49.8 |
| KK3 Max, Bluetooth | 265.18 | 18.3 | 51.3 |
| KK3 Pro, cable | 884.17 | 3.00 | 34.2 |
| KK3 Pro, 2.4G dongle | 856.9 | 19.6 | 43.0 |
| KK3 Pro, Bluetooth | 144.84 | 21.5 | 57.8 |
| KK3, cable (GPDL) | 144.36 | 7.95 | - |
| KK3, Bluetooth (GPDL) | 88.17 | 24.9 | - |
| KingKong 2 Pro, cable | 248.76 | 18.0 | 25.3 |
| Elves 2 Pro, cable | 893.66 | 5.34 | 22.35 |
| Elves 2 Pro, Bluetooth | 87.29 | 25.24 | 55.05 |
| ES Pro, cable | 978.42 | 1.75 | 3.45 |
| ES Pro, Bluetooth | 717.68 | 3.53 | 3.70 |
| Elves 2, cable | 992.2 | 1.85 | 4.20 |
| Elves 2, Bluetooth | 712.42 | 3.52 | 5.13 |
| ES, cable | 999 | 1.98 | 4.01 |
| ES, Hyperlink 2 adapter | 969.93 | 3.76 | 4.73 |
| ES, Bluetooth | 799.36 | 4.00 | 4.44 |
| ES Max, cable | 999 | 1.88 | 2.75 |
| ES Max, Hyperlink 2 adapter | 900.9 | 5.26 | 5.74 |
| ES Max, Bluetooth | 796.81 | 4.19 | 5.02 |
| TT Pro, cable | 996.02 | 2.55 | 1.93 |
| TT Pro, Hyperlink 2 adapter | 984.25 | 4.59 | 3.97 |
| TT Pro, Bluetooth | 798.72 | 3.04 | 2.24 |
| TT Max, cable | 1001 | 2.54 | 2.03 |
| TT Max, Hyperlink 2 adapter | 981.35 | 3.51 | 3.54 |
| TT Max, Bluetooth | 799.36 | 3.22 | 2.41 |
Test details: KK3 Max cable, adapter and Bluetooth, firmware 5.7, 21 November 2025; KK3 Max dongle, firmware 5.6, 16 June 2025. KK3 Pro, firmware 5.1, 19 December 2024 (gamepadla shows the page as "User Verification" only). KK3, 27 December 2024. KingKong 2 Pro, firmware 5.1, 6 July 2026. Elves 2 Pro, firmware 5.7, 9 October 2025. ES Pro and Elves 2, firmware 2.7, August 2025. ES, firmware 3.8, 1 April 2026. ES Max, firmware 4.2, 11 August 2026. TT Pro, firmware 6.2, 25 January 2026. TT Max, firmware 5.8, 14 and 15 January 2026. The twelfth GuliKit pad on gamepadla, the discontinued NS08, is left out: it is Bluetooth-only, and its primary test, a button run from 2023, read 48.01 Hz and 71.49 ms.
The KK3 Max dongle: 1000 Hz and 30 ms slower
A single pair of tests could be a bad day. This is not one. Gamepadla's extra results include four sessions where a tester measured the KK3 Max cable and dongle on the same firmware, and the gap barely moves:
| Session | Cable (ms, Hz) | Dongle (ms, Hz) | Dongle costs (ms) |
|---|---|---|---|
| Violet Prismriver, firmware 4.1, 30 April 2024 | 3.06 at 1000 | 33.22 at 1000 | 30.16 |
| Scott Cerato, firmware 4.1, 3 and 4 May 2024 | 3.05 at 1000 | 29.14 at 1000 | 26.09 |
| John Punch, firmware 5.0, 19 November 2024 | 2.70 at 859.85 | 32.82 at 853.14 | 30.12 |
| John Punch, firmware 5.6, 16 June 2025 | 3.80 at 887.95 | 35.39 at 884.46 | 31.59 |
Three testers, three firmware generations, 26 to 32 ms every time. At 884.46 Hz the mean wait for the next report is 0.57 ms, from the half-interval rule on our polling rate page (derived). Polling explains under 2% of the gap. The rest is the wireless path, and the KK3 Pro shows the same pattern at a smaller size: 3.00 ms on the cable and 19.6 ms on its dongle, 16.64 ms apart. GameSir's worst dongle on our GameSir polling rate page cost 23.87 ms; the KK3 Max is worse than that in every session.
Sticks are slow on both links. The KK3 Max measured 33.9 ms on the cable and the KK3 Pro 34.2 ms, at close to 1000 Hz. The ES Pro, on the same kind of cable, measured 3.45 ms.
Why reviewers say the dongle is just as fast
The 1000 Hz figure is real, and reviewers measured it. Neowin's review (Robbie Khan, April 2024) lists the polling rate as "Variable up to 1000Hz wired/wireless (2.4GHz only)" and says latency "remains excellent whether 2.4GHz wireless or wired. This makes sense given that the dongle offers a 1000Hz polling rate which means a 1ms latency can be expected." HLPLANET found the cable and dongle "close to 1000Hz" in its tests.
Neowin's numbers came from software, "Gamepadla and Xinput Tester", which count how often reports arrive, and the review itself notes that "only a dedicated piece of hardware sitting between controller and PC can measure accurate input latency." Gamepadla's Prometheus 82 rig is that hardware, and gamepadla's own page makes the point: a high polling rate "does not automatically guarantee lower latency if the controller's internal processing is slow." The KK3 Max dongle is the clearest case of it on this site: a polling rate within 11% of the cable, and latency 6.6 times higher (35.4 against 5.40 ms, derived).
Hyperlink 2: Bluetooth at 800 Hz
GuliKit's ES and ES Pro launch release claims "Bluetooth polling rates of 730Hz / 3.25ms lowest latency and wired rates of 1000Hz / 1.87ms lowest latency". Its TT Series release gives 730 Hz "when connected directly via Bluetooth on PC", 820 Hz through the Hyperlink 2 adapter and 1000 Hz wired, all measured with GPDL. Those are best-case figures, and gamepadla's averages land close: 712.42 to 799.36 Hz over Bluetooth on six pads, with buttons between 3.04 and 4.19 ms. Bluetooth cost 0.49 to 2.31 ms over the cable on each pad (derived).
That breaks the rule our 8BitDo Ultimate polling rate page draws from 8BitDo's pads, where every model fell to 66 to 200 Hz over Windows Bluetooth. That range belongs to those controllers, not to Windows. The pad's firmware sets the Bluetooth rate: a Microsoft Xbox pad measured 133 Hz on our Xbox controller polling rate page, a DualShock 4 728 Hz. GuliKit's own older pads show it: the KK3, KK3 Pro and Elves 2 Pro, without Hyperlink 2, measured 87 to 145 Hz.
The adapter is not a latency upgrade on these pads. It raised the rate to 900.9 to 984.25 Hz, but buttons were slower than built-in Bluetooth on the ES Max, TT Pro and TT Max, and 0.24 ms faster on the ES (derived). GuliKit's TT release says it now ships the adapter "to reduce interference caused by inconsistent Bluetooth receivers, 2.4GHz Wi-Fi congestion, or shared bandwidth", which is a reason to use it on a PC whose Bluetooth misbehaves. On a KK3 Max it does little: 249.19 Hz and 18.3 ms, the same latency as the pad's own Bluetooth.
Firmware moved the KK3 Max Bluetooth
The KK3 Max's Bluetooth changed more between firmware versions than anything else in its data. On firmware 5.6 in June 2025 it measured 103.93 Hz and 29.06 ms on buttons; on 5.7 in November, same tester, 265.18 Hz and 18.3 ms. Neowin called the Bluetooth "trash" on PC in 2024, and HLPLANET measured it "below 125Hz". On firmware 5.7 gamepadla measured more than double that rate, with buttons still more than three times the cable's 5.40 ms.
Where this stops applying
Each row is one unit tested by one person, and gamepadla gives Prometheus 82 a margin of plus or minus 1 ms, so gaps under a millisecond are inside it. The KK3 Max primary dongle test is on older firmware than its cable test; the session table above is the like-for-like comparison. Bluetooth results depend on the PC's own radio, which GuliKit itself names as the reason for its adapter. Switch mode on a cable caps every GuliKit pad tested at about 125 Hz, the same penalty our Switch Pro Controller polling rate page finds on other brands. And one GuliKit claim does not survive our data: its Gamescom 2025 release says a DualSense gets "only 125Hz polling rate in regular Bluetooth mode on PC", while gamepadla measured 542.42 and 798.72 Hz, as our DualSense polling rate page shows.
What to do
- KK3 Max or KK3 Pro on PC: use the cable. The dongle polls at nearly the same rate and measured 17 to 32 ms slower on buttons.
- ES, ES Pro, ES Max, Elves 2, TT Pro or TT Max: plain Bluetooth is fine, at about 800 Hz and 3 to 4 ms. The cable still wins by up to 2.31 ms.
- Use the Hyperlink 2 adapter only if your PC's Bluetooth drops or stutters. On four pads it beat built-in Bluetooth once, by 0.24 ms.
- Buying a GuliKit for PC, choose a Hyperlink 2 model. The Elves 2 Pro, KK3 Max and KK3 Pro measured 22 to 34 ms on wired sticks.
- Keep the pad in XInput. Switch mode drops every model to about 125 Hz.
Frequently asked questions
Is the GuliKit KK3 Max 1000 Hz?
On the cable, yes: gamepadla measured 992.06 Hz under Windows 11 on firmware 5.7, with buttons at 5.40 ms. On the bundled 2.4G dongle, in four like-for-like sessions, it measured 853 to 1000 Hz, but buttons averaged 29.14 to 35.39 ms. The polling rate holds; the latency does not.
Is the KK3 Max dongle slower than the cable?
Yes, by about 30 ms on buttons. In four sessions where gamepadla testers measured the cable and the dongle on the same firmware, the dongle cost 26.09 to 31.59 ms, at nearly the same polling rate. Use the cable on PC.
Is GuliKit Bluetooth bad on PC?
It depends on the model, and here is the split. The ES, ES Pro, ES Max, Elves 2, TT Pro and TT Max use GuliKit's Hyperlink 2 Bluetooth and measured 712 to 799 Hz and 3.04 to 4.19 ms on buttons over Windows Bluetooth. The KK3, KK3 Pro, KK3 Max and Elves 2 Pro measured 87 to 265 Hz and 18.3 to 25.2 ms.
Do I need the Hyperlink 2 adapter?
Not for latency on a PC whose Bluetooth works well. On the ES, ES Max, TT Pro and TT Max, the adapter raised the rate to 900 to 984 Hz, but buttons were slower than built-in Bluetooth on three of the four, by up to 1.55 ms, and 0.24 ms faster on the ES. GuliKit says it includes the adapter to avoid inconsistent Bluetooth receivers and 2.4 GHz congestion. On a KK3 Max it measured 249 Hz and 18.3 ms.
Is the Elves 2 Pro the same as the Elves 2?
Not on PC. The Elves 2 measured 712.42 Hz and 3.52 ms over Bluetooth, the Elves 2 Pro 87.29 Hz and 25.24 ms. Wired, the Elves 2 Pro's sticks took 22.35 ms against 4.20 ms on the Elves 2.