Quiet Keyboards for Open Offices and Video Calls

A quiet mechanical keyboard for office work depends on three things: the switch, what dampens the case, and whether the keycaps are PBT or ABS. Most boards marketed as silent use rubber-dampened switches, typically adding ~10 grams of actuation force but cutting the clack on both press and release. Case foam and desk mats absorb the rest.

The microphone hears more than the person across the room does. Noise suppression helps, but it compresses your voice along with the typing.

Key points

• Linear switches with dampening rings (silent reds, silent blacks) measure quieter than tactile equivalents by 5-8 dB at typical typing speed.
• Case foam and a desk mat absorb the resonance that travels through the frame and the surface; the switch alone is not the whole problem.
• PBT keycaps produce a lower, duller sound than ABS; the frequency matters more for microphone pickup than the raw decibel reading.
• Software noise suppression (Krisp, RTX Voice, Discord's built-in gate) cuts typing bleed but also flattens dynamic range in speech.
• Low-profile boards typically measure 3-6 mm shorter travel and register as quieter on calls because there is less distance for the key to bottom out.
• A five-minute recording with your actual microphone in your actual position will tell you more than any manufacturer's claim.

Why keyboards end up audible on calls: mic pickup, not room volume

Your microphone sits closer to the keyboard than to your mouth. A cardioid pattern rejects sound from behind but not from directly below, which is where the board sits. The desk itself acts as a sounding board: every key press sends a vibration through the frame and into the surface. Your mic hears that thud before it hears the click.

Noise gates help. They mute the channel when you are not speaking, so the typing between sentences disappears. Suppression algorithms (Krisp, RTX Voice, the one built into Zoom) go further and try to separate voice from everything else in real time. Both work. Both also compress what is left.

The problem with relying on software is that it costs dynamic range. Your voice gets flatter, sibilants blur, and if you are explaining something technical with varied intonation the algorithm smooths it into a monotone. You are better off with a board that does not need aggressive gating in the first place.

What makes a board quiet: switch choice, case foam, keycap material

Switches determine most of the sound. A standard Cherry MX Red bottoms out with a sharp plastic clack. A Cherry MX Silent Red uses rubber bumpers top and bottom to absorb the impact at both ends of travel. You lose a few milliseconds of response, you add weight to the press, and you cut the noise by enough that it stops bleeding through a noise gate.

Silent linears (silent reds, silent blacks, silent pinks) measure quieter than silent tactiles because there is no additional event mid-press. A tactile bump adds a second frequency. That is fine in a room on your own. It is not fine when someone is trying to listen to you through a condenser microphone two feet away.

Case construction matters almost as much. A hollow plastic shell rings when a key bottoms out. Foam inside the case damps that resonance. Some boards ship with it; others leave a cavity you can fill yourself with neoprene or sorbothane. The difference is audible even on silent switches.

Keycap material changes the frequency more than the volume. ABS is lighter and produces a higher-pitched clack. PBT is denser and sounds duller. Your ears might not care, but a microphone's diaphragm picks up high frequencies more readily, so PBT caps often read as quieter on a recording even when a decibel meter shows similar levels.

A desk mat absorbs the final piece: the vibration that travels from the board into the desk and back up into the mic stand. Neoprene or thick felt under the keyboard is worth more than another round of switch replacements.

The quiet boards we stock

We stock four keyboards chosen for calls and shared spaces. Each uses a different approach. None will click through a noise gate if you type at normal speed.

Keychron K5 Version 2

The Keychron K5 uses low-profile Gateron switches with 3 mm of travel instead of the usual 4 mm. Red or brown. The shorter throw means less impact at the bottom, and low-profile boards sit flatter on the desk so the typing angle is shallower. The frame is aluminium but hollow; adding foam to the cavity cuts the ring noticeably.

It connects over Bluetooth or USB-C. Battery lasts about a week with the backlight off. The keycaps are ABS, so if you want the duller sound of PBT you will need to replace them — low-profile caps are less common than standard, so check fitment.

Good for someone who types all day and wants the feel of a mechanical switch without the noise penalty of full travel. Not silent, but quiet enough that it disappears behind a noise gate.

Logitech MX Mechanical

Logitech's tactile quiet switches use rubber dampening top and bottom. The press is heavier than an undampened tactile — around 65 grams peak instead of 55 — but the sound is muffled enough that it does not register through Zoom's built-in suppression. The linear option is quieter still; the clicky option defeats the point.

The case has an internal dampening layer. The keycaps are ABS but thick enough that they do not sound thin. It connects over Bluetooth, Logitech's Bolt USB receiver, or USB-C wired. You can pair it to three devices and switch with a key. Battery lasts months.

This is the one to get if you do not want to think about foam, mats or keycap swaps. It is quiet out of the box and it stays that way.

AULA F99

The AULA F99 ships with silent red switches and both foam and silicone dampening in the case. It is a 100-key layout with a numpad, so it is wider than a tenkeyless but still fits a standard desk. The frame is plastic, which normally rings, but the dampening stops it.

It connects over 2.4 GHz wireless, Bluetooth, or wired USB-C. The wireless response is fast enough that you will not notice latency in normal typing. Keycaps are PBT. The silent reds feel slightly mushy compared to undampened linears, but that is the trade-off: less feedback, less noise.

Good if you need a numpad and you do not want to spend Logitech money. The silent reds are effective and the PBT caps help.

Cherry KW X ULP 2.0 MINI

Cherry's ultra-low-profile switch measures 3.5 mm total travel. The MX ULP Click variant is marketed as clicky, which sounds wrong in a quiet keyboard roundup, but the click is internal to the switch housing and quieter than a standard blue by a wide margin. The low-profile chassis does not resonate the way a taller case does.

It is a compact 65% layout, so no function row and no numpad. Connects over Bluetooth or 2.4 GHz wireless. Battery lasts months. The keycaps are low-profile PBT.

This works if you are used to a laptop keyboard and you want mechanical feedback without the height or the noise. The ultra-low-profile feel is closer to a scissor-switch MacBook keyboard than to a standard mechanical, which is either ideal or a dealbreaker depending on what you are coming from.

Software fixes: noise suppression and what it does to your voice

Krisp, RTX Voice, and the suppression built into Zoom, Meet and Teams all use trained models to separate speech from other sounds. They work. They also compress your voice into a narrower frequency band, which flattens your tone and blurs consonants slightly. If you are on calls for an hour a day that is fine. If you are explaining a technical problem for three hours the listener will notice the monotone.

Noise gates are simpler and less destructive. The microphone mutes when you stop speaking, so the typing between sentences disappears but your voice when you are talking stays untouched. Discord, OBS and most DAWs let you set the threshold manually. Set it just above the noise floor of your room and below your quietest speech. Test it by typing while speaking and adjusting until the typing does not reopen the gate.

The best fix is a microphone with a tight cardioid or hypercardioid pattern, positioned close and angled slightly away from the keyboard. A dynamic mic (SM58, Procaster, PodMic) rejects off-axis sound better than a condenser, so the keyboard has to be much louder to bleed through. That costs more and needs an interface, but it is the permanent solution if you are on calls daily.

You can also just ask. If someone tells you they hear typing, believe them and fix it rather than assuming your noise gate is good enough.

The five-minute test before you join a call on a new board

Record yourself. Open Audacity, Zoom, or whatever you use for calls. Start recording, speak a few sentences at normal volume, then type for ten seconds at normal speed. Stop recording and play it back.

If the typing is as loud as your voice, the board is not going to work on calls without aggressive suppression. If the typing is audible but quieter, a noise gate will handle it. If the typing disappears into the noise floor, you are fine.

Do this with your actual microphone in your actual position. A board that is quiet enough for a headset boom mic two inches from your mouth might still bleed through a desk-mounted condenser a foot away. The geometry matters more than the switch spec.

Test it again with noise suppression turned on, then turn suppression off and listen to how your voice sounds in both states. If the suppressed version sounds flat or muffled, you have lost something. Decide whether the trade-off is worth it or whether you need a quieter board instead.

Common questions

Are silent switches slower for typing or gaming?

The rubber dampening adds 1-2 milliseconds to the return on release. You will not notice it in typing. In gaming, fast double-taps feel slightly less crisp than undampened linears, but the difference is small enough that most people adapt within a few hours. If you are playing at a level where 2 ms matters, you already know whether you need an undampened switch.

Can I make a loud keyboard quieter with keycaps alone?

Swapping ABS for PBT will change the frequency and make it sound less sharp, but it will not cut the volume by enough to matter on a call. The switch and the case are the bigger problems. If the board has a hollow case, add foam. If the switches are not dampened, replace them or accept that software suppression is doing the work.

Do low-profile keyboards feel worse than standard height?

They feel different. Less travel, less key wobble, flatter typing angle. Some people adapt immediately; others find the shorter throw unsatisfying and go back to standard height within a week. The only way to know is to try one. Low-profile is quieter on average because there is less distance to bottom out, but that does not make it better if you hate typing on it.

Will a desk mat really make a difference?

Yes. A thick neoprene or felt mat under the keyboard absorbs the vibration that would otherwise travel through the desk and into the microphone. The improvement is smaller than switching to silent switches, but it is not nothing. If you have a condenser mic on a desk arm or a boom stand, the mat is worth the money.

What about O-rings on the keycaps?

O-rings sit on the stem under each keycap and cushion the bottom-out. They help a bit — maybe 3-5 dB — but they also make the key feel mushy at the bottom of travel and they do nothing for the sound on release. Silent switches with built-in dampening are more effective because they dampen both ends. O-rings are cheap, though, so if you already have the board and you do not want to replace switches they are worth trying.

How quiet is quiet enough for an open office?

If your coworkers can hear it from six feet away over ambient office noise, it is too loud. Silent switches with a desk mat will usually get you there. If you are in a very quiet space — a library, a recording studio, someone's living room — you might need a membrane keyboard instead. Mechanical switches all make some noise; the question is whether that noise is acceptable in context.

Can I use a mechanical keyboard on calls if I use push-to-talk?

Yes. Push-to-talk mutes you by default, so the typing never reaches the microphone unless you are holding the key. This is the simplest solution if you are gaming or in a voice channel where you do not need to be live all the time. It does not work for meetings where you are expected to respond immediately.

Which is quieter: silent reds or silent blacks?

Silent blacks are slightly heavier — around 60 grams actuation instead of 45 — but the noise level is nearly identical because both use the same dampening design. The difference in force means you are less likely to bottom out hard on blacks, which can make them sound marginally quieter in practice. Try both if you can; most people choose based on feel rather than noise.

What to get

The Logitech MX Mechanical with tactile quiet switches is the least effort for the most reliability. It is quiet out of the box, it connects to three devices, and it does not need foam or a keycap swap to work on calls. Get it if you want a quiet mechanical keyboard and you do not want to modify anything.

The Keychron K5 Version 2 with low-profile reds is quieter than standard-height boards and cheaper than the Logitech, but it needs foam in the case to stop the frame ringing. Get it if you are comfortable opening the case and adding dampening yourself, or if you prefer the feel of low-profile switches.

The AULA F99 is the option with a numpad. Silent reds, PBT caps, foam and silicone dampening all included. It is not as refined as the Logitech but it works and it costs less. Get it if you need the numpad and you do not want to spend flagship money.

Do not get any of these if you are gaming at a competitive level and you care about response feel over noise. Silent switches are slower on release and the dampening makes fast double-taps feel slightly dead. In that case, get an undampened linear board and rely on push-to-talk or a dynamic microphone instead.