How HMX took over the mid-tier linear slot
Two years ago HMX was a name on a vendor page nobody pronounced confidently. Today the Cloud sits second on our W19 tracker and lands in half the $120 prebuilts the scene argues about. The rise wasn't an accident.
The pass-5 Trends Tracker currently has HMX Cloud Linear at +36 for W19 — second only to Gateron Oil King and well clear of every Cherry, Gateron, and Tecsee SKU underneath. Two years ago HMX was a name on the back of a CannonKeys product page, often misspelled as "HMK" in the threads that mentioned it at all. The brand is now the default factory-lubed linear in the $0.45–$0.55 bracket, ships in roughly half the prebuilts the 75-percent-default crowd argues about, and is the answer to "what should I drop into my Mode Sonnet R2" almost as often as the Oil King is. The interesting question isn't whether HMX is good. It's why a brand that was a curiosity in 2023 is the load-bearing supplier in the mid-tier slot in 2026 — and why neither the legacy incumbent nor the established challenger could keep that slot from changing hands.
What HMX is, briefly
HMX is the switch arm of an injection-molding house out of Huizhou that started shipping under its own name in mid-2023. Before that the same factory was running OEM contracts for several of the names the scene already knew — the open secret in the b-roll of every early HMX review is that the housing tooling had been working since well before the brand existed. That matters, because HMX did not arrive with prototype-grade tolerances. The first Cloud, Hyacinth, and Macchiato trays out of the gate measured cleaner stem wobble than the equivalent Gateron Pro 2.0 trays from the same window, and reviewers — Theremin Goat first, then half of YouTube within a quarter — picked up on it.
The release cadence since has been disciplined. HMX shipped roughly one new SKU every six to eight weeks across 2024 and 2025, broadly clustered in three families: light POK-stem linears (Cloud, Pinball, Macchiato), mid-weight POM-stem linears (Sunset Gleam, Hyacinth), and a small bench of tactiles that haven't materially threatened the Tecsee Sapphire's hold on that slot. The product page on hmxkeyboard.com reads like an older Gateron catalog circa 2021 — short copy, a force-curve graph, materials called out plainly, no narrative. That restraint is part of why the brand reads as serious. It's also part of why, when builders type on the switches and the switches sound like the marketing claimed, the trust compounds quickly.
The factory-lube competence
The single decision that put HMX on the bench is that the factory lube on a Cloud out of the bag is, to a very near approximation, what a careful builder would lay down by hand with GPL 105 on the rails and a thin film of Krytox 205g0 on the stem legs. It is not a coat. It is not a slick. It is a film, applied where friction lives, and absent from the leaf and the housing wells where it would mute the switch. The first time a builder pops a top off a stock Cloud and looks at the rail under a desk lamp, the lube pattern is almost suspiciously neat.
What this collapses is the buy-build-use loop time. The 2021-vintage modding workflow was: order switches, wait, open switches, lube switches over an evening with a bent paperclip and a 5cc of 205g0, wait an hour for the lube to settle, install, type, hate one of them, pull it, redo it. The 2026 workflow with HMX is: order switches, install switches, type. That delta is not visible on a spec sheet but it is enormous in the lived experience of the hobby — and crucially, it is the delta a first-time builder feels most strongly. A person buying their second keyboard, who knows enough to read about lube but not enough to enjoy the four-hour pass, gets a switch that already sounds like the reviews said it would. The reviews stop carrying the asterisk that was on every linear recommendation between 2018 and 2022: "after lube."
The downstream effect is that HMX gets credit for sound that, on a force-curve and a materials-list, ought to read as ordinary. A 42g/50g linear in a nylon-over-nylon housing with an LY stem is not, on paper, a switch that should be a flagship. The Cloud is a flagship because it arrives ready to type, and the seventy switches in a tray sound consistent across the tray. Both are factory discipline, not material novelty.
POK stems and the mid-tier sound profile
The Cloud's stem is LY — a nylon-blend formulation HMX uses on its lighter linears — and the bench is split fairly evenly with the brand's POK-stem releases (the Macchiato, the Hyacinth in some batches). Either way the stem sits in a lower-friction class than the POM stems that dominated the 2022–2023 mid-tier and a different acoustic class than the UHMWPE stems on the boutique tactile bench. POK in particular, where HMX uses it, gives the switch a distinctive top-out: brighter than POM, drier than UHMWPE, with a little woody crack to the upstroke that builders chasing a poppy 75-percent sound have come to specifically request.
Where this lands the Cloud is in a register the scene didn't have a clean name for in 2024 and now reflexively calls "clack-forward." The bottom-out is contained — nylon-on-nylon, modest weight, no ring. The top-out carries the character. Compared against the Oil King's deep, fused thock, the Cloud sounds like a different instrument: separated events, faster decay, a higher fundamental. Compared against the Cherry MX2A Red's firm-conservative tune, the Cloud sounds livelier and a half-step lighter under the hand. The point is not that one of these voicings is correct. The point is that HMX picked one — articulated, light, top-end forward — and shipped it consistently across an entire SKU family, in a year when the rest of the mid-tier was still negotiating between clones of the Oil King's voicing and clones of the older Cherry-derived voicing.
A switch family with a clear acoustic identity is easier to recommend, easier to film, and easier for a vendor to merchandise. Half of the HMX-equipped builds in the build-of-the-week rotation across late 2025 are recognizably HMX from the audio alone. That kind of audio fingerprint is brand equity that compounds.
The price tier HMX defended
Read the order pages and the discipline gets clearer. HMX prices its standard linear bench in the $0.45 to $0.55 per-switch range at major North American vendors — Cloud at $0.45 on CannonKeys at the time of writing, Hyacinth at $0.50, Macchiato at $0.55. The brand has not chased the $0.20 utility-tier slot that the Akko CS line and the Outemu V3 housings occupy, and it has not chased the $1.00+ enthusiast tier where Tecsee, Geon, and the small-batch boutique makers operate. Every public price across the catalog sits inside a 25-cent band.
That band is not arbitrary. It's the band a $100–$150 prebuilt can absorb. A Keychron Q1, a Mode Sonnet R2 at the kit-and-switches tier, a YUNZII at the upper end of its pricing — all of these can put HMX switches into a $130 final BOM and still hit margin. The same builds cannot absorb $1.00 Tecsees without breaking the price ceiling, and they cannot absorb $0.20 utility switches without breaking the trust the prebuilt category needs from its switch supplier. HMX is the switch you can put in a prebuilt and not have to apologize for in the listing.
The implication for the brand is that HMX wins the volume tier by not competing for the slots above and below it. The implication for the field is that this slot — the "good factory-lubed linear in a serious prebuilt" slot — was held by Gateron from roughly 2020 to 2023, and is held by HMX now, and was never really held by Cherry at all. The MX2A revision was Cherry's attempt to participate at this tier and the timing was wrong. The slot was occupied while the revision was still in tooling.
Restock cadence — the quiet advantage
The argument that gets the least attention and is doing the most work is restock cadence. HMX restocks at the major North American vendors on roughly an eight-week cycle — not perfectly, and not on every SKU, but reliably enough that a Cloud out-of-stock notice rarely runs longer than two months. Gateron's cadence on the Oil King and Pro 3.0 line runs closer to quarterly, with longer gaps on the boutique-finish variants. Cherry's cadence is functionally infinite — the MX2A is a production-line product and is always available — but the trade-off is that the product is also never fresh; there is no reason for a vendor to feature it because it never arrives.
Eight weeks turns out to be the right number. It's tight enough that vendors can carry HMX as a default category — every build guide can specify "HMX Cloud or equivalent" without a footnote about availability — and it's loose enough that each restock generates a small amount of news. A new tray hitting CannonKeys or NovelKeys is a Discord-pingable event in a way that a Cherry restock is not. The brand gets the freshness premium of Gateron's quarterly drops and the carry-them-as-default reliability of Cherry's production line, without paying the cost of either model in full.
This is a compounding advantage. A vendor whose default linear restocks every two months will write build guides around it. A vendor who writes build guides around it will sell more of those builds. Those builds will appear in the build-of-the-week rotation and on review channels. The reviews will drive the next eight-week restock to sell through. None of this works at twelve weeks. None of this works at four weeks either — at four weeks the news cadence is too fast to organize merchandising around. Eight weeks is the sweet spot, and HMX is the brand currently hitting it.
Why the structural-incumbent disadvantage is the real story
The temptation in a piece like this is to frame HMX's rise as a story about plucky-newcomer-beats-sleeping-incumbent. That isn't what happened. The structural problem the incumbents face is that the slot HMX won is a slot that requires a mid-sized, focused supplier. Cherry is too large; its production economics push it toward office-prebuilt and OEM contracts where the volumes justify the tooling. Gateron is the right size but its catalog is too broad; the Oil King is a flagship that consumes the brand's acoustic identity and pulls attention away from the mid-tier line behind it. Tecsee is too small and too premium; the brand earns its margin on the Sapphire and the boutique tactiles, not on volume linears.
HMX is the size that the mid-tier slot rewards: large enough to restock on cadence, small enough to ship a coherent acoustic identity across the whole catalog, focused enough on switches alone that the brand does not have to balance a keycap line or a board line or a vendor channel. That focus is the sibling argument to the one we ran in the MX2A retreat piece: Cherry's load-bearing role in the hobby moved to keycap manufacturing because that's where the brand's scale is the asset rather than the constraint. HMX is on the other side of that same equation. Mid-tier switches reward focus. HMX is focused.
What we're watching
Three threads. The first is whether HMX ships a tactile that genuinely contests the Sapphire's hold on the boutique-tactile bench. The brand has not yet, and the catalog gap is conspicuous; the rumor inside the b-stock channels is that an HMX tactile rework is in tooling for the back half of 2026. If it lands and is competent, HMX moves from "the mid-tier linear brand" to "the mid-tier brand."
The second is the wireless and Hall-effect crossover. HMX is a mechanical-switch maker working in a moment when a measurable share of new prebuilts are shipping with magnetic switches and analog input. The brand has not signaled a Hall-effect SKU and is on record, in the Theremin Goat interview from late 2025, as skeptical of the category. That position will be tested across 2026 as Wooting-style boards push past the gaming tier into the typist tier.
The third is whether the eight-week restock cadence holds. The cadence is the quiet engine under the brand and the easiest thing to lose; a single supply-chain disruption or a single decision to chase a higher-margin tier could break the rhythm. If HMX is still restocking the Cloud on roughly its current cadence at the end of 2026, the brand has converted what looks like a moment into a position. If it isn't, the slot opens again, and the next supplier sized correctly for the slot — whoever that turns out to be — gets the same opening HMX got in 2024.
Until either of those threads tightens, the read is steady. HMX took the slot because the slot rewards a focused mid-sized supplier shipping factory-lubed competence on an eight-week clock. The brand is currently the only one doing all three things at once. That is most of the explanation, and most of why the +36 on the W19 tracker is not, on inspection, surprising.
Keep reading
- Deep Dives
What Cherry's MX2A revision actually changed
Cherry's MX2A landed in 2023 with a real engineering update — factory lube, tighter tolerances, the same 100M-press lifespan. The reason the discourse cooled isn't that the switch failed; it's that the enthusiast slot it aimed at was already claimed by the time it arrived.
thock7 min read - Deep Dives
Why the Gateron Oil King sounds the way it does
Few stock linears earn the word 'creamy' on first press. The Oil King does, and the reason is not a secret stem material or a proprietary lube — it is a careful pairing of housing resins and a factory tune that other linears under thirty cents tend to skip.
thock5 min read - Deep Dives
Why the Drop Holy Panda X feels the way it does
The Holy Panda was a six-month frankenswitch project for most of its life. The Drop Holy Panda X is the first version where the geometry, the spring tune, the housing pairing, and the factory lube were the spec — not the mod.
thock12 min read - Deep Dives
Why clicky switches still have a constituency
Linears and tactiles dominate the modern build conversation, and clickies barely appear in it. The reason is not that clicky switches have aged out — it is that the people who want them really want them, and the manufacturers know it.
thock9 min read