What people actually mean by ’65 keyboard’ on an ergonomic desk
When people search for a 65% keyboard, they're usually thinking in mechanical-keyboard terms: that means keeping your arrow keys but ditching the function row and numpad. In the ergonomic space, that exact formula almost never shows up—what you actually get are a few compact split designs that tackle the same goal differently, mainly by removing the numpad and shrinking the overall footprint so you're not stretching your arms all over the place. If you showed up wanting a small, angled, numpad-free keyboard setup, the options on this list represent what you can realistically grab from Amazon today.
Compact and split are two different decisions
Going compact (dropping the numpad) and going split (separating the two halves) are independent choices, and it’s worth picking them separately instead of assuming one implies the other. The 60% keyboard and 75% keyboard pages on this hub cover the pure size question without the split requirement. This page leans toward boards that do both, because that combination is what most people mean when ‘65%’ and ‘ergonomic’ show up in the same search. If you only want smaller, not split, a tenkeyless board or a mini keyboard is a shorter trip with less adaptation time.
Tenting matters more than the key count
Once you’ve decided you want a split, the number that actually changes how your wrists sit is the tenting angle, not whether the board technically qualifies as 65% or 75%. The Freestyle2 with VIP3 lifters gives you 5, 10, or 15 degrees; the Goldtouch Elite’s ball joints go all the way to 30. A flat split with no tenting option — like the bare Freestyle2 — is still a real improvement over a rigid one-piece board, but it’s a smaller improvement than the same frame with lifters attached. Decide whether you actually want tenting before you decide how compact you want to go.
Wired vs. wireless is not a minor detail here
Across this category, wireless split boards carry a failure mode that wired ones simply don’t have: the link between the two halves can drop, on top of the usual link to your computer. That shows up in Amazon’s own review summaries often enough that I treat it as a pattern, not a fluke. If you’re buying a compact split specifically to use daily at a desk that doesn’t move, wired removes one entire category of future annoyance. Save wireless for boards you’re actually carrying between rooms or desks.
What I’d skip, and why
I'd pump the brakes on anything here with under 50 reviews—that includes the Freestyle Edge RGB Plus, the Goldtouch Elite, and definitely the Truly Ergonomic CLEAVE. When a keyboard has a stellar rating from just a handful of reviewers, that's not proof of quality yet; it's early data that hasn't exposed the weak spots. I should also mention that the split keyboards people actually rave about most—the ZSA Moonlander, the MoErgo Glove80, Dygma's Defy—aren't even available through Amazon. You have to buy them straight from the companies that make them. There's no point pretending otherwise to pad out a list; if you're hunting for the most acclaimed split keyboard in existence, you'll find it on those manufacturers' websites, not here.
Pairing a compact split with the rest of your desk
A smaller keyboard changes your hand position, but it doesn’t fix a desk that’s the wrong height or a chair that isn’t supporting you. Check your desk height and seating before assuming the keyboard is the whole answer, and pair whichever board you pick with a proper wrist rest — it’s the cheapest, best-reviewed intervention in this entire hub. If you’re also rethinking your pointing device to match a split layout, the ergonomic mice page covers that half of the desk, and the Alice layout page is worth a look if you want a single-piece split board instead of two separate halves.
How I picked these
I worked from a verified research brief of Amazon listings — prices, star ratings, and review counts read directly off live product pages, not estimated or carried over from memory. I did not test any of these boards myself; there’s no lab here, no measured wrist angles, no typing-speed benchmarks. Where a listing’s review count was too small to mean anything, I said so instead of pretending the star rating settles the question. Where a product genuinely isn’t sold on Amazon, I left it out of the ranking and told you where it actually lives instead.
