At-home IPL, compared on energy, cooling and skin-tone safety

Intense pulsed light works by being absorbed preferentially by the melanin in a hair follicle, heating it enough to disrupt growth. That mechanism is also the whole safety story: the device cannot tell the difference between melanin in a hair and melanin in surrounding skin. It only works when there is a strong contrast between the two.

Every home IPL device sold in the United States has a skin-tone ceiling, usually expressed against the Fitzpatrick scale. Above that ceiling the device is not merely less effective, it is a burn risk, because the surrounding skin absorbs the pulse instead of the follicle. Brands state this, but typically in a compatibility chart three clicks deep rather than on the box.

They are equally quiet about the other end. Very light blonde, red, white and gray hair contains too little melanin to absorb the pulse at all. No home IPL device works on it, at any energy level, on any skin tone. If that is your hair color, no device on this page will do what you want, and the honest answer is electrolysis.

This is the most important page on this site and it is not a product page. Read the skin tone safety guide first. Then come back and compare energy.

Joules per square centimeter is the figure to compare, where it is published. The spread is enormous. Ulike states 23.6 J/cm2 for the Air 10. Braun states up to 6 J/cm2 for the Silk-expert Pro 5. Nood states 5 J/cm2 for The Flasher 2.0. Philips publishes no J/cm2 figure for the Lumea line at all.

A four-fold difference in stated energy does not mean four times the result, and higher is not automatically better: more energy means more heat at the skin surface, which is exactly why Ulike pairs its figure with sapphire contact cooling held near 16 to 18 degrees Celsius. Braun takes the opposite approach, running lower energy with a sensor that reads your skin eighty times a second and sets the intensity for you. Both are defensible designs. One asks you to tolerate heat for fewer sessions; the other asks you to accept more sessions for a gentler pulse.

Home IPL hair removal sits under FDA product code OHT. The records here are in better shape than in the LED category.

Ulike's UI20 family, which includes the Air 10, is covered by K241998 (December 2024), with the UI20S variants under K250942 (July 2025). The Braun Silk-expert Pro 5 is K190354 and the Pro 3 is K190366, both cleared 18 March 2019 under applicant Cyden Limited. Philips Lumea holds current records K243453 (February 2025) and K253754 (December 2025). JOVS hand-held IPL devices hold K220645 and K214113.

Nood returns nothing. We searched the public database by brand and by device name and found no record for The Flasher 2.0, while some retail listings for it say "FDA approved", which is a phrase that does not apply to any device in this category. Cleared and approved are different things, and a retailer that confuses them is not a reassuring sign.

• Contact cooling is the difference between a treatment you complete and one you abandon. Sapphire-cooled handsets hold the window near 16 to 18 degrees, which lets you tolerate the higher energy levels that make the course shorter.
• Flash count matters less than it sounds. Ulike quotes 1,000,000 flashes, Philips 450,000, Braun 400,000, Nood around 600,000. All of these exceed a full-body course many times over. SmoothSkin's unlimited-flash models remove the question entirely.
• Treatment window size is a real trade-off. A wide window covers legs fast and is clumsy on the upper lip. Philips sells separate attachments for body, face, bikini and underarm precisely because one window does not do all four well.
• The schedule is the hard part. Most protocols run weekly for four to twelve weeks, then maintenance. Missing weeks does not just delay results, it means treating hairs out of phase with their growth cycle. Our treatment schedule guide explains why.

Start with the main roundup, or go straight to the best options for deeper skin tones if that is your question. It is the question most of this category answers badly.