Diode vs Alexandrite vs Nd:YAG: Which Hair Removal Machine to Buy in 2026

Most buyers comparing laser hair removal machines start with output power, then brand, then price. That order is backwards.

Wavelength decides which clients you can treat at all. Power decides how fast you treat them. Price decides how long it takes to earn the machine back. Get the wavelength architecture wrong and no amount of added wattage rescues the decision — you will simply be treating a narrower slice of your local market, more slowly, with a machine you cannot resell easily.

Three wavelengths dominate professional hair removal: 755nm, 808nm and 1064nm. Almost every buyer eventually asks the same question — diode versus alexandrite versus Nd:YAG — and the honest answer is that two of those three names describe a crystal, while a diode platform can now reproduce both of their wavelengths inside a single handpiece. That changes the economics of the decision, but it does not make the trade-offs disappear.

A note on scope: this guide is written for the person buying the machine — a salon owner, a clinic manager, a distributor adding a SKU. It covers wavelength architecture, real specifications, throughput and what to verify before you pay. It is not about what a single session costs a patient; if that is what you need, start with how many sessions diode laser hair removal actually needs and come back here when you are ready to specify hardware.

The quick answer, by business situation

Skip to your row. The reasoning behind each answer is in the sections that follow.

Your situation	Build that fits	なぜ
First machine, single treatment room, mostly Fitzpatrick I–III clients	Single-wavelength 808nm or 810nm diode	Lowest purchase cost, smallest footprint, no unused capability sitting in the cabinet. The 808–810nm band covers the largest share of a typical walk-in client base.
Mixed client base across I–IV, one cabinet has to cover the whole menu	Triple-wavelength 755 + 808 + 1064nm diode	Three clinical profiles from one handpiece. You can treat light-skin fine hair and darker-skin coarse hair without turning anyone away or buying a second unit.
High volume, light-skin market, speed is the constraint	Dedicated alexandrite (755nm)	A diode handpiece sharing three wavelengths cannot match a dedicated 755nm source on peak power, and peak power is what buys treatment speed on large areas.
Predominantly Fitzpatrick V–VI, or a large volume of darker skin types	Platform that includes a genuine 1064nm option at usable fluence	1064nm is the only one of the three with a workable safety margin on deeply pigmented epidermis. This is the one requirement you should not compromise on.
Distributor or multi-unit buyer	Triple-wavelength diode	Widest resale fit — a single SKU that can be sold into almost any clinic profile, rather than a specialist unit that only clears in light-skin markets.

What each wavelength actually does to a follicle

All three wavelengths work through selective photothermolysis: the beam is absorbed by melanin in the hair shaft and follicle, converted to heat, and the follicle is destroyed by thermal damage while the surrounding tissue is spared. What separates the three is how strongly melanin absorbs at each wavelength, and how deep the light travels before it is scattered away.

Melanin absorption falls as wavelength rises. Penetration depth rises as wavelength rises. Those two curves moving in opposite directions create the entire trade-off — and they create it three times over.

755nm: the fastest melanin hit

755nm has the strongest melanin absorption of the three. That means more heat generated per unit of energy delivered, which translates into effective treatment of fine and lighter hair that other wavelengths struggle to catch. It also means the shallowest penetration — the energy is largely spent in the upper dermis, where the follicle sits in thin areas but where the epidermis also competes for that energy. On light skin this is efficient. On deeply pigmented skin the epidermis absorbs a disproportionate share, and the risk of superficial injury climbs quickly.

808nm: the workhorse

808nm sits at the balance point. Melanin absorption is lower than at 755nm, penetration is deeper, and the ratio between follicular damage and epidermal heating is favourable across a wide band of skin types. This is why 808nm became the default diode wavelength for professional hair removal and why the 808/810nm single-wavelength build remains the most common first machine. It does almost everything reasonably well and nothing spectacularly — which is exactly what a general-purpose clinic wants.

1064nm: the safety margin on darker skin

1064nm has the weakest melanin absorption and the deepest penetration. Less energy is absorbed in the epidermis, so the surface stays cooler and the margin for error on Fitzpatrick IV–VI opens up. The cost is efficiency: with less absorption per unit of energy, you need higher fluence and more passes to reach the same follicular temperature, and very fine light hair may not respond at all. 1064nm is not a “stronger” setting — it is a different compromise, and it is the only one of the three that makes darker skin types clinically practical.

波長	Melanin absorption	Penetration	Best skin types	Where it earns its keep	Honest limitation
755nm	Highest	最も浅い	I～III	Fine and light hair, large light-skin areas, fast body work	High epidermal risk on IV–VI; narrower safe operating window
808nm / 810nm	中程度	Moderate to deep	I–IV	General body work, mixed client base, the default professional setting	On very dark skin the epidermal margin narrows; on very fine light hair it is slower than 755nm
1064nm	Lowest	最深部	IV–VI	Deeply pigmented skin at higher fluence, coarse dark hair, deeper follicles	Needs higher fluence and more passes; poor response on fine light hair; slower per area

Skin type comes before brand

Every adverse event story in this category traces back to the same root cause: a wavelength was pushed onto a skin type it was not suited to, usually because the assessment was wrong or rushed. The machine cannot compensate for a misclassification, and a more expensive machine will not fix it either.

Before you choose hardware, be honest about the distribution of Fitzpatrick types in your catchment area. A clinic in a Northern European city and a clinic in a Mediterranean or Middle Eastern city can be running identical business models and need completely different wavelength priorities.

Fitzpatrick type	Typical profile	Preferred wavelength	What to watch
I	Very fair, burns easily, never tans	755nm or 808nm	Low fluence to start; pigment changes are the main risk, not burns
II	Fair, burns easily, tans minimally	755nm or 808nm	Well covered by a single 808nm build
III	Light brown, sometimes burns, tans gradually	808nm, 755nm with care	The upper end of comfortable for 755nm; test patch strongly advised
IV	Moderate brown, rarely burns, tans easily	808nm, 1064nm	755nm is no longer a safe default; cooling quality starts to matter more than wavelength
V	Dark brown, very rarely burns	1064nm	Epidermal heating dominates; longer pulse widths and aggressive cooling are essential
VI	Deeply pigmented, never burns	1064nm	Only 1064nm with a real cooling system; expect more sessions and lower clearance per pass

If more than roughly a fifth of your client base sits at IV or above, a 1064nm option is not an optional upgrade — it is the difference between being able to serve that fifth of the market and having to refer them out.

Why “triple wavelength” exists — and what it actually is

Here is the part most spec sheets blur. In an alexandrite laser, 755nm comes from alexandrite crystal rod. In an Nd:YAG laser, 1064nm comes from a neodymium-doped yttrium aluminium garnet rod. In a triple-wavelength diode platform, 755nm and 1064nm are produced by diode stacks tuned to emit at those wavelengths — not by a crystal rod of either type.

That is a real engineering approach, not a marketing fiction. It has genuine advantages, and it has a genuine cost. Both matter to your purchase decision.

What you gain. One handpiece, three clinical profiles, no crystal rods or flashlamps to service, no consumable per shot, instant switching between wavelengths mid-appointment, and a single cabinet instead of two. For a clinic that treats a mixed client base in one room, this is the only architecture that covers the whole menu without buying twice.

What you give up. Peak power per wavelength. A handpiece rated at 1,000W that shares its output across three wavelengths is not delivering 1,000W at 755nm — that energy is divided. A dedicated alexandrite concentrates its entire output at 755nm, and that concentration is exactly what buys speed on large light-skin areas. The gap is real and measurable in treatment time, not just on paper.

Where a dedicated alexandrite still wins outright: high-volume light-skin body work, very fine or light-coloured hair, and any setting where the business model is built on short appointment slots and maximum clients per day. If you are running twenty full-leg sessions a day on Fitzpatrick I–III clients, no diode platform will match a true alexandrite on throughput. The two-way comparison — and the cases where diode wins on comfort and running cost — is set out in diode versus alexandrite in practice.

Where a dedicated Nd:YAG still wins outright: deeply pigmented skin at high fluence, and any clinic that also wants 1064nm for non-hair work. A diode stack tuned to 1064nm is a practical hair-removal solution but it is not the same instrument as a Q-switched or long-pulsed Nd:YAG. For the standalone comparison, see Nd:YAG versus diode for hair removal.

Where the triple wins: the ordinary single-site clinic with a mixed book. It is the pragmatic answer for most buyers, which is why it has become the default configuration in this price band.

Four builds, one factory, side by side

The comparison above is abstract until you put real numbers next to it. These four configurations come from the same production line, which makes them a useful controlled comparison — the differences you see are engineering choices, not different factories.

仕様	EN191H	EN137D	EN218	EN221B
波長	755 + 808 + 1064nm	755 + 808 + 1064nm	810nm single	Diode 808 + 755 + 1064nm, plus Nd:YAG handle
ハンドル	1	1	1	3 (diode, Nd:YAG, AI analyser)
Interchangeable heads	5 (10×10, 13×20, 13×30, 13×40mm + nose-hair head)	Fixed 15×25mm	Fixed 12×18mm	5 diode heads (15×15, 15×20, 15×30, 15×40mm, 46mm) + 5 Nd:YAG heads
エネルギー密度	Up to 150 J/cm²	1–150 J/cm²	0–50 J/cm²	1–100 J/cm²
頻度	Up to 10Hz	1-10Hz	1-10Hz	-
ハンドル・パワー	1,000W	500W (1,000W / 1,200W optional)	-	500W Nd:YAG handle · 3,500W machine
Shot life	~20 million	-	50 million	30–50 million
冷却	4 cooling levels	-	Air + semiconductor + TEC	Sapphire crystal (diode), water + air + TEC
Working modes	shrスタック、shr、hr	-	-	-
ディスプレイ	15.6インチ	15.6インチ	15.6-inch touch	15.6-inch Android HD touch
Machine weight	35kg packed	20kg · 31.5kg packed	52.6kg packed	75kg packed

Read the table as four different businesses rather than four grades of quality. EN137D is a single-handle triple at moderate power for mixed skin types — the entry point. EN191H adds interchangeable heads, a 1,000W handle and in-motion modes, which is what turns a general-purpose unit into a high-throughput one: four spot sizes mean you stop compromising between small areas and large ones. EN218 is the opposite bet — one wavelength, 50 million shots, heavy cooling, built for continuous general body work where the client base is predominantly light to medium skin. EN221B stacks a hair-removal platform, a tattoo removal handle and a skin analyser into one cabinet, which is a different product category entirely: it is for clinics that want to open three service lines from one purchase.

Each of the four is documented on its own page if you want the full datasheet rather than the comparison view: EN191H, the 5-head triple-wavelength diode; EN137D, the single-handle triple; EN218, the 810nm single-wavelength workhorse; and EN221B, the combined diode and Nd:YAG platform. If your decision is really between two wavelengths inside the diode family rather than three across technologies, the narrower comparison is covered in 808nm versus triple-wavelength diode builds.

Six specifications that matter more than the wavelength count

Once the wavelength architecture is settled, these six numbers decide whether the machine performs in daily use. Most buyers compare only the first two.

1. Spot size. Bigger spots treat area faster and push energy deeper, because less of the beam is lost to scattering at the edges. But a larger spot needs more total energy per pulse to hold the same fluence, which is why big spot sizes and high fluence rarely coexist on a low-power handpiece. A 13×30mm or 13×40mm head is a genuine throughput advantage for legs and backs; a 10×10mm head is for face and small areas. If a supplier offers only one fixed spot, check that it matches the areas you actually treat most.

2. Fluence ceiling. The maximum energy density determines your headroom on coarse hair and darker skin. 50 J/cm² is workable for a light-skin general practice; 100–150 J/cm² gives you room to escalate on resistant hair and to run 1064nm at clinically useful levels. A ceiling you never approach is marketing. A ceiling you hit in month two on your most common client is a problem.

3. Repetition rate and in-motion mode. Up to 10Hz matters far more than it sounds. In-motion delivery (often labelled SHR) lets an operator glide the handpiece continuously instead of stamping, which is faster, more comfortable, and far less operator-dependent. If a machine offers only stamp mode at low frequency, appointment times will be longer than the spec sheet implies.

4. Cooling. Cooling quality decides both comfort and the safe fluence ceiling, especially on Fitzpatrick IV and above. Sapphire contact cooling held at low temperature, or air plus semiconductor plus TEC, are the two credible approaches. Ask what the crystal surface temperature actually reaches and how long it holds that temperature under continuous use — not what the brochure claims.

5. Pulse width. Pulse width has to sit between the thermal relaxation time of the epidermis and that of the follicle: long enough to spare the surface, short enough to damage the target. Adjustability across roughly 10–400ms covers the useful range for hair removal. A fixed pulse width locks you into one skin type.

6. Handle power and shot life. Handle power drives how much energy reaches the skin after losses; shot life drives the replacement cycle. A 20 million shot handpiece and a 50 million shot handpiece are not equivalent purchases — the second one is a lower cost per treatment over its life, even at a higher initial price. Ask for the replacement handpiece price before you sign, not after.

The throughput math nobody shows you

Purchase price is the number buyers negotiate over. Revenue per treatment room hour is the number that determines whether the machine was a good decision. The two are only loosely connected.

Capacity works out as usable hours multiplied by sessions per hour, and sessions per hour is driven almost entirely by spot size, repetition rate and whether the machine can run in motion.

Configuration	Typical spot	Full-leg session time	Legs per hour	Legs per week (30 treatment hours)
Single 808nm, 12×18mm, stamp mode	2.16 cm²	~35–45 min	1.3–1.7	40–50
Triple, fixed 15×25mm, in-motion	3.75 cm²	~25–30 min	2.0–2.4	60–72
Triple, 13×40mm head, in-motion	5.20 cm²	~18–22 min	2.7–3.3	80–100

Put a session price on those rows and the picture changes fast. At a market-typical €40 per full-leg session, the difference between the first row and the third is roughly 40 extra sessions a week — a five-figure annual revenue gap from two machines that may sit within 30% of each other on price. This is the argument for paying for spot size and in-motion delivery rather than for the highest wattage on the brochure.

Payback follows directly:

Payback period = machine cost ÷ (sessions per week × price per session × gross margin per session).

Run it with your own numbers before you compare quotes. Two machines at different prices frequently land on the same payback month, and occasionally the more expensive one wins — because capacity, not price, is the binding constraint on a treatment room.

Consumables, spares and the five-year cost

Diode platforms have one structural advantage over older flashlamp technology: there is no lamp to replace and no per-shot consumable. Gel, disposables and eye protection are the only recurring items, and they are trivial against any of these machine prices. That does not mean the five-year cost is low — it means the cost moved somewhere less visible.

It moved into the handpiece. A handpiece rated for 20 million shots sounds unlimited until you do the arithmetic: a full-leg session consumes roughly 1,500–2,000 pulses with a large spot at in-motion settings. Twenty million shots is therefore roughly 10,000–13,000 full-leg sessions — around five to six years at eight legs a day, 250 days a year, and considerably less if you run high-frequency small-area work. A 50 million shot handpiece roughly doubles that window.

The other three costs to model are cooling maintenance (water quality and filter changes on water-cooled units), downtime during service, and voltage/plug configuration for your market. None of them appear on a quotation, and all of them show up in year three.

Eight checks before you transfer money

Specifications on a page are claims. These eight checks convert claims into verified facts, and every one of them can be done before you pay a balance.

Check	How to verify it	Failure signal
1. Actual output wavelength	Ask for a spectrometer reading per wavelength, or verify independently on arrival	Refusal to provide any measurement; “same as everyone else” as an answer
2. Real spot size	Fire one pulse onto thermal paper or a coated card and measure the burn mark	Measured area clearly below the quoted rectangle
3. Cooling performance	Measure the crystal surface with an infrared thermometer, then again after 20 minutes of continuous use	Temperature climbs steadily and never recovers
4. Power stability under load	Run at maximum frequency for 30 continuous minutes and watch for output droop or a thermal cut-out	Machine throttles down within minutes; cooling loop undersized
5. Shot counter integrity	Confirm the counter is readable and ask whether it can be reset — and whether that voids the handpiece warranty	Counter absent, or reset capability sold as a feature
6. Spare handpiece economics	Get the replacement handpiece price and lead time in writing, with the quote	Price quoted only after the sale; long lead time with no local stock
7. Voltage and plug configuration	Confirm the unit ships configured for your market’s voltage and plug type	External step-up transformer offered as a solution
8. Training and documentation	Confirm operator training language, a written treatment protocol per skin type, and a service contact that answers	Protocols verbal only; no named service contact

For a deeper version of items 6 to 8, including how to audit a supplier’s production depth rather than their showroom, see our 12-point guide to vetting a beauty equipment manufacturer.

Matching the build to your service menu

The last filter is not technical. It is what you intend to sell.

Business type	ビルド	Reasoning
Neighbourhood salon adding hair removal to an existing menu	Single 808/810nm diode, fixed mid-size spot	One wavelength covers the bulk of walk-in demand at the lowest entry cost; upgrade later when volume justifies it
Skin clinic with a mixed, urban client base	Triple 755 + 808 + 1064nm, interchangeable heads	Serve every skin type in one room without referring anyone out; multiple spots cover face through back
High-volume body-work studio in a light-skin market	Dedicated alexandrite, or triple with a large spot head	Throughput is the constraint; alexandrite for maximum speed, large-spot triple for flexibility
Clinic serving predominantly Fitzpatrick V–VI	Platform with genuine 1064nm capability and strong cooling	Safety margin is the deciding factor, not speed; do not accept a 755nm-dominant build
Clinic wanting hair removal plus tattoo removal or skin analysis	Combined platform (diode + Nd:YAG + analyser)	Three service lines, one cabinet, one service relationship

One further fork is worth naming, because it changes the budget more than any specification on this page: laser versus intense pulsed light. IPL sits below all three wavelengths discussed here on both selectivity and speed, but it also sits well below them on price, and for a low-volume salon treating mostly light skin it is sometimes the rational entry point. The trade-offs are laid out in diode versus IPL — read it before you assume laser is the only option at your volume.

Frequently asked questions

These are the six questions that come up in almost every enquiry, answered directly.

Is 755nm in a diode machine the same as an alexandrite laser?

The wavelength is the same; the source is not. 755nm from a diode stack produces the same melanin absorption profile as 755nm from an alexandrite rod, so the clinical logic for skin type and hair colour is identical. What differs is peak power per wavelength — a dedicated alexandrite concentrates its full output at 755nm, while a triple-wavelength diode divides its output across three.

Can a 1064nm diode really treat Fitzpatrick V–VI safely?

It can, provided the machine reaches usable fluence at 1064nm and the cooling system holds temperature under continuous load. The wavelength gives you the margin; the cooling and the operator’s fluence discipline determine whether that margin is actually used. Test patch, conservative starting fluence and longer pulse widths remain non-negotiable.

Do I need three wavelengths, or is 808nm enough?

If your client base sits at Fitzpatrick I–III, 808nm alone is enough and buying three wavelengths is paying for capability you will not use. If more than about a fifth of your clients are type IV or above, the 1064nm option stops being optional — those clients either get treated properly or they go elsewhere.

How many shots should a handpiece be rated for?

Treat 20 million as the practical floor for professional use and 30–50 million as the comfortable range. Below 20 million, the replacement cycle arrives while the machine is still new to you, which distorts the cost per treatment badly.

What spot size should I insist on?

Match it to your dominant treatment area, not to the largest number on the sheet. Legs and backs want 13×30mm or larger; face and small areas want 10×10mm to 13×20mm. Interchangeable heads let you hold both, which is the single most useful upgrade on a general-purpose machine.

Can one machine handle both hair removal and tattoo removal?

Yes, in a combined platform that pairs a diode hair-removal handle with a separate Nd:YAG or picosecond handle. They are separate handpieces with separate physics — the diode does the hair, the second handle does the ink. If tattoo removal is a serious service line rather than an add-on, size it separately; the tattoo removal machine price guide covers the wavelength and payback logic for that side of the business.

The bottom line

Wavelength architecture is the decision. 755nm buys you speed on light skin, 808nm buys you general-purpose coverage, 1064nm buys you the right to treat darker skin types at all. A triple-wavelength diode gives you all three in one handpiece and accepts lower peak power per wavelength in exchange. A dedicated alexandrite or Nd:YAG gives you maximum performance in one narrow lane and leaves the rest of the market to someone else.

For most single-site buyers with a mixed book, the triple is the correct answer — and within that answer, spend the difference on spot size, in-motion delivery and cooling rather than on wattage.

If you want the wavelength question settled against your actual client mix rather than a generic recommendation, tell us the Fitzpatrick distribution you see and the areas you treat most, and we will map it to a specific configuration — including the entry-level and high-throughput builds on the same platform. Send us your requirement and we will come back with the specification, not a brochure.