How to Measure Switch Force: gf, cN, and the Numbers on the Box (2026)
Every switch product page has a force number on it. Almost nobody explains what that number actually measures — and the gap between “45” on the box and what your finger experiences is the distance between three different points on a curve that the box never shows. This guide walks through the vocabulary (gf vs cN, actuation vs bottom-out), the real shape of a force curve, and what you can honestly measure at home — using official specs from Cherry’s MX2A datasheet and Keychron’s Super switch line as the worked examples.
This is the companion piece to our switch types explainer — that article covers how switch families behave; this one covers what the numbers mean.
The three force points that matter
A switch’s spring pushes back harder the further the stem travels. “Force” on a spec sheet is therefore never one number — it’s a curve, and manufacturers quote individual points on it:
| Term | What it measures | Where it sits on the curve |
|---|---|---|
| Initial force | Force holding the stem at rest, before it starts moving | Travel = 0 mm |
| Actuation (operating) force | Force at the moment the keystroke registers | Usually 2.0–2.2 mm down |
| Bottom-out (final) force | Force at full travel, stem pressed against housing | Full travel (4.0 mm on standard MX) |
The number the community argues about is actuation force — it decides how easily a key registers. The number your fingertips actually feel at the end of every keystroke is bottom-out force. Both belong on the spec sheet; most brands print only one.
gf vs cN: why the units don’t match between brands
Cherry, the original MX manufacturer, specs its switches in centinewtons (cN). Almost everyone else — Keychron, Gateron, Gazzew, and the AliExpress brands — uses gram-force (gf). The conversion: 1 cN = 1 / 0.980665 gf ≈ 1.02 gf (a centinewton is a defined SI-derived unit; gram-force depends on standard gravity, which is why the two differ slightly).
So a Cherry MX2A Red at “45 cN” actuation is 45.9 gf — about 2% heavier than a “45 gf” Keychron Super Red. Small enough that forum threads treat them as equal; precise enough that force-curve databases don’t. When you see a comparison table mixing the two units, check which column is which before concluding one switch is “lighter.”
Worked example: two official spec sheets side by side
Cherry publishes the fullest public datasheet in the industry, including the initial and final force points almost no other brand quotes. Keychron specs in gf and adds tolerance bands. Both below are quoted from the official product pages:
| Spec | Cherry MX2A Red (datasheet, cN) | Keychron Super Red (product page, gf) |
|---|---|---|
| Type | Linear | Linear |
| Actuation / operating force | 45 cN (~45.9 gf) | 45±15 gf |
| Initial force (rest) | 30 cN (~30.6 gf) | not published |
| Final force (bottom-out) | 100 cN (~102.0 gf) | not published |
| Pre-travel | 2.0 mm | 2.0±0.4 mm |
| Total travel | 4.0 mm | 4.0±0.4 mm |
| Tactile sibling | MX2A Brown (tactile, no click) | Super Brown (55±15 gf), Super Banana (57±8 gf, earlier bump) |
| Lifespan | >100 million keystrokes (gold crosspoint) | 50 million cycles |
Two things in this table are worth pausing on:
- Cherry’s “final force 100 cN” is not bottom-out weight you’d hold. It’s the spring resistance at the very bottom of the stroke on the spec point grid — the figure counts the full compressive load at 4.0 mm travel, which includes what the housing takes over from your finger at bottom-out. Independent force-curve measurements of Cherry Reds land well below that number at practical bottom-out. The lesson isn’t “Cherry lies” — it’s that final force is the least standardized number on any spec sheet, and comparing final forces across brands is mostly noise.
- Keychron’s ±15 gf tolerance band is honest and huge. A “45 gf” Super Red can measure anywhere from 30 to 60 gf out of the box. Double-stage springs narrow the variation across the stroke (Keychron’s own wording: the double-stage long spring “reduces the tolerance of spring force”), but unit-to-unit spread of budget springs is real. This is one more reason a single number never describes a switch.
What a force curve actually looks like
A force curve plots force (Y) against travel (X). The shapes that matter:
- Linear (Cherry MX2A Red, Keychron Super Red): a smooth upward ramp from initial force to bottom-out. No kinks. The finger feels a spring getting progressively stiffer, nothing else.
- Tactile (MX2A Brown, Super Brown/Banana): the ramp rises to a bump peak before actuation, dips slightly as the stem passes the bump (“snap-over”), then ramps to bottom-out. A “stronger tactile” switch is one with a taller bump relative to the dip — which is exactly what Keychron claims for the Super Banana (“earlier tactile bumps and stronger tactile feedback than a regular brown switch”, per the official page).
- Clicky (MX2A Blue): same mechanism as tactile plus a click jacket that produces the audible click at the bump. The click and the tactile event are the same physical moment.
The curve is also why two switches with identical actuation force can feel nothing alike: same point on the Y axis, different shape getting there. Cherry can quote 45 cN and Gateron can quote 45 gf and the curves can still cross twice between 0 and 4 mm.
How to measure it yourself (honestly)
Three realistic levels, cheapest first:
| Method | What you get | What it can’t do | Cost |
|---|---|---|---|
| Switch tester | Relative feel of 4–10 switches, same keycap, same board | Any numbers at all | $10–25 |
| Kitchen scale (0.1 g) | Approximate bottom-out force per switch | The curve shape; actuation point needs registering the switch mid-press | $0 (most kitchens) |
| Force-curve rig (load cell + travel encoder) | The real curve, the only valid basis for comparisons | — be realistic: this is lab gear | $100+ DIY, much more for proper rigs |
Kitchen-scale method, briefly: place the bare switch (or the whole board) on the scale, tare, then press straight down with something rigid and flat — a pen cap, a single keycap — and read the peak value the scale flashes. That peak is roughly bottom-out force. It’s crude (angle, speed and scale sampling all add error), so treat ±10% as the margin. If your “45 gf” switch measures 38 or 52, both are inside the printed tolerance.
For real curves, the community reference is independent measurement — Haata’s force curves database is the standard citation — because a curve drawn from a marketing sheet is an illustration, not a measurement.
So which force do you actually want?
Sensible defaults, consistent with the official specs above:
- Long typing sessions, fatigue-sensitive: light linear around 45 gf (MX2A Red, Super Red). Cherry’s own positioning calls Red “the first choice for beginners.”
- Feedback for typing accuracy: medium tactile around 55–58 gf (Super Brown 55±15 gf, Super Banana 57±8 gf, MX2A Brown).
- Gaming: actuation happens at 2.0 mm / ~45 gf on every mainstream switch — faster than any in-game event requires. If you bottom out and mistype in games, a slightly heavier spring fixes more than a “speed” label does.
- Shared/office space: force won’t make a switch quiet; that’s a dampening question. See switch types explained for silent variants.
The one rule that survives every argument in this niche: before buying 70+ switches, buy one tester or one board with hot-swap and a set of switches — the force you like is an empirical fact about your hands, not a spec-sheet conclusion.