How I Make Vintage-Style Custom Keycaps
Most people starting with custom keycaps gravitate toward modern designs — clean lines, bold colors, low profiles. I found myself looking the other way. There is something about the slightly worn, imperfect aesthetic of old typewriter and computer keys that feels right for certain setups. Getting there is not as complicated as some guides make it sound, but there are practical traps if you do not know what to look for. The process I use centers on three main paths: modifying existing keycaps with surface treatment, commissioning a small batch from a manufacturer, or 3D printing your own. Each has tradeoffs that become obvious only after you have ruined a set trying the wrong approach. I will walk through my actual workflow, starting with the method I use most often. Surface modification on stock keycaps.
I buy blank ABS or PBT keycap sets — usually from reputable brands like GMK or Varmilo blanks — and then treat them. The vintage look I am after comes from a combination of color fading, texture changes, and sometimes slight dimensional alterations. The main technique involves a light sanding with 800-grit paper followed by a thin coat of matte clear sealant. This gives the plastic a slightly roughened, aged surface that catches light differently than factory-fresh caps. Here is where beginners go wrong. They sand too aggressively and destroy the legends before they even start the coloring phase. I sand in light, circular passes over maybe thirty seconds per key, checking frequently by running my thumb across the surface. You want the shine to go away, not the detail. Once the gloss is removed, I apply the sealant in two very thin coats with a spray can held at least eight inches away. One thick coat will pool around the stem and make the keycap physically unusable on most switches. For color, I do not paint the whole keycap. I use a dry-brushing technique with acrylic paint diluted with a little water. The brush gets barely any paint on it, then I drag it lightly across the raised surfaces. This leaves pigment only on the high points, which mimics natural wear patterns. A genuine 1980s IBM Model M keycap does not have uniform color — the edges and tops are faded from decades of finger contact while the recesses stay darker. Replicating that requires restraint.
I ran into a specific problem with this method a while back that took me a few attempts to solve. I was working on a set of Cherry Profile keycaps and discovered that the dry-brush technique worked beautifully on the top surface but left the legend area completely bare because the paint had no grip on the smooth, slightly concave legend wells. My keys looked like they had been sanded clean through the lettering. The workaround was straightforward once I figured it out: I applied a very light first coat of paint directly into the legend areas using a fine brush, let it dry for ten minutes, and then dry-brushed the surrounding surface as usual. This layered approach preserved the legends while still giving the aged look I wanted. It adds about twenty minutes to the process per keycap set, which is noticeable but acceptable. The second path — commissioning a small batch — works best if you want consistent results across an entire set without putting in hours of manual labor. I have worked with a few manufacturers on custom runs. The key specification to communicate clearly is the profile. Vintage typewriter keys used a much deeper dish shape than modern Cherry or OEM profiles. If you order "vintage style" without specifying the profile, you will get a modern cap with a weird color treatment, which looks worse than doing nothing at all. I always request a physical sample before committing to a full run. One bad sample costs less than a full order of misplaced expectations. The third option is 3D printing. This gives you the most control over shape and size, but the surface finish out of the printer is not immediately vintage-looking. I print in resin at 50-micron layer height, then do the same sanding and sealing process I described above. Resin prints absorb sealant differently than injection-molded plastic, so I test on a spare keycap first. A sealant that works perfectly on PBT will fog or soften a resin print if you are not careful about the formulation. I use a water-based polyurethane sealer for resin and a solvent-based one for ABS blanks. Mixing those up ruined a full set of prints once. I do not recommend learning that lesson the hard way.
There are some counter-intuitive things about vintage keycap design that most tutorials skip. First, age is not uniform. A real vintage keycap has wear concentrated on the areas your fingers actually touch — the upper outer edges for the home row, the flat top for spacebars and large modifiers. If you treat the entire keycap evenly, it looks artificially aged rather than genuinely worn. I focus my dry-brushing and light sanding on these specific contact zones and leave the areas between your fingers relatively untouched. Second, the stem matters more than people realize. Vintage keycaps from the 1970s and 1980s often had thicker, sturdier stems with different tolerances than modern Cherry MX-compatible stems. If you are building a restoration project for an actual vintage keyboard, modern keycap stems may be too loose or too tight in the original switches. I keep a small collection of period-correct keycaps from donated or discarded vintage keyboards specifically for stem compatibility reference. This has saved me from buying entire sets that would not physically fit the boards I was restoring. A downside of the surface modification approach is that it is entirely destructive to the original keycap. Once you sand and seal, you cannot reverse it. If you mess up a set, that set is gone. This is why the spare keycap testing step is non-negotiable. Always run through the entire process on one or two spare caps before committing to the full set.
Another limitation: surface-treated keycaps will continue to wear over time, just like the originals they emulate. The sealant layer will eventually break down at high-contact points, which some people consider part of the charm and others consider a failure of the project. If you want something that stays consistent without periodic re-treatment, commissioning from a manufacturer using industry-grade dye-sublimation or double-shot molding is the more durable route, though it costs significantly more per set. I also want to mention a tool I use that is probably overkill for most people but has made the process more reliable for me: a digital caliper. Vintage keycaps vary in height and dish depth depending on the manufacturer and era. Measuring before and after modification lets you track exactly how much material you are removing and whether the final profile still works with your intended switches. It takes about five minutes per keycap but prevents a lot of guesswork. If you are just starting out, I would suggest beginning with a single modified keycap on an existing board rather than committing to a full set. Put one vintage-treated key on your spacebar or Enter key and type on it for a few days. See how the texture feels, how the legend readability holds up, and whether the aesthetic actually improves the experience for you. This alone tells you more than any forum thread or video tutorial, and it prevents buying materials for a full project you end up hating.
The materials I typically use cost between fifteen and forty dollars depending on how many keycaps you are treating. A basic sanding kit, a couple of spray sealants, some acrylic paint, and a set of blank keycaps. The time investment is the real variable — a full set done carefully takes me roughly three to four hours spread across multiple sessions to allow drying time between steps. That is the practical overview. The specific techniques take practice, the materials are inexpensive, and the results are genuinely satisfying if you have the patience to do them right.
