Product1 distinct publisher3 min readPublished
The CameraJet's endoscope reads 28 frames a second and fires rinse into the gaps it finds. The headline 69 percent plaque number comes from Dyson's own lab, in one brushing mode, on plaque Dyson helped invent.
The Product Desk · Product desk

product
Dyson hands the flossing decision to a camera inside the toothbrush1 distinct publisher
product
Dyson charges $500 for a toothbrush that aims its own floss jet1 distinct publisher
product
Dyson bets $499.99 on a toothbrush that aims its own floss jet1 distinct publisher
product
One suspension, four students, and an AI image nobody in the chain could delete1 distinct publisher
Compiled by The Product DeskSomething wrong?How this is made
At the sink, none of this is visible to the person holding the brush. The model gets about 2.8 frames to work with: 28 images a second, and 100 milliseconds between the camera seeing a gap and the jet flushing it [2][4][17]. A macro lens is something any competitor can buy. The 470,000 dental images Dyson says it collected over six years of development, roughly 78,000 a year, are the part a rival would have to spend six years acquiring [3][18].
Then there is the number that will end up on the box. Dyson says the CameraJet removes up to 69 percent more plaque than market-leading premium electric toothbrushes in hard-to-reach areas, and its own small print puts that in laboratory testing, in brush-only variable mode, in those hard-to-reach areas [5][6]. The test substrate is proxy plaque Dyson developed with the National University of Singapore's Faculty of Dentistry, a method the company adopted so designers could iterate without a clinical trial after every minor change [7]. That is sensible engineering practice and weak consumer evidence, because it is a comparison Dyson designed, ran, and scored.
Read "brush-only" carefully. If the jet was off for that test, the flagship figure is testimony about bristle contouring and variable sonic oscillation rather than about the camera at all [6][15]. Dyson does not spell that out either way, so that is my reading of the caveat, not the company's statement.
The most honest line in the whole announcement is that you do not need to open the MyDyson app to use the toothbrush [12]. Coverage mapping and guided cleaning exist in there [16], and Dyson has already priced in how often people will actually tap them by making the hardware work with the app shut. The value case cannot rest on app sessions, then; it rests on those 100 milliseconds, the only part that happens whether or not anyone opens anything.
Dyson also says the camera is live only during app view or automatic scanning, and that images are not recorded or stored on the device or in the cloud [13]. That is the right answer for a camera in a bathroom. It also means the installed base never teaches the model anything after launch, so those 470,000 images are the whole curriculum for the life of the product [3][13].
For anyone shipping a sensor into a commodity category, two things decide whether it justifies a premium: whether the sensor changes what happens in the moment or only reports on it afterward, and whether the buyer can verify the benefit themselves or only in the vendor's lab. Dyson has built a genuine closed loop on the first count and a lab claim on the second [4][6]. Plaque in the gaps between molars is not something a customer can see in week one, so the premium is being charged against trust in a test rig. If your product sits in that quadrant, the next thing to build is one outcome the customer can measure without you. WIRED notes this is not Dyson's first hybrid, after the short-lived Zone air-purifying headphones, though it calls the pairing far more practical this time [14]. Whether it is proven is still a separate question.
Ranked by verification strength, evidence, and original report placement.
Dyson announced the CameraJet, a two-in-one electric toothbrush and flosser, on Tuesday.
A built-in endoscope camera in the CameraJet analyzes 28 images per second, shooting bursts of rinse as the user moves the device over their teeth.
Dyson calls the system Gap Optical Targeting and says it is powered by a macro-lens camera and a machine learning algorithm trained on 470,000 dental images collected during the six years it spent developing the product.
Dyson says the camera sees the gap and flushes it with mouth rinse within 100 milliseconds, while the user brushes normally.
Dyson's small print says the 69 percent plaque figure is based on laboratory tests in the brush-only variable mode, in hard-to-reach areas.
Dyson created its own proxy plaque in collaboration with the National University of Singapore's Faculty of Dentistry, allowing designers to iterate and test plaque removal in the lab without running a clinical trial after every minor change.
Distinct publishers with included, body-backed reporting in this cluster.
1 article · September 1, 2026
Follow any of these and your For You feed starts watching them — no settings page required.
Evidence-backed comparisons of source perspectives and observed adoption signals. Read the methodology
Which Builder, Operator, and Investor concerns the observed source mix emphasized—not a truth score.
Evidence, demonstrated adoption, hype gap, incentives, and confidence are assessed independently, each on its own current evidence. How these are measured.
One briefing, printed carefully
WIRED is thorough on mechanism and, to its credit, reproduces the qualifier under the 69 percent figure rather than burying it. But thoroughness about a company's claims is not measurement of them: there is no clinical trial, no reviewer time, no teardown, and no second publisher to compare against. The single number Dyson leads with is fenced by three conditions in Dyson's own footnote.
Announced, priced, unused
A price, two colours and a dock are not uptake. This is launch day: nobody in this reporting has brushed with the thing for a fortnight, no dentist is quoted, no ship date or unit figure appears. The only behavioural number offered — roughly one in three US adults flossing daily — describes the gap Dyson wants to sell into, not any traction it has.
The number outruns the test behind it
The overstatement is not in the invention, which sounds real, but in a chain where every link belongs to Dyson: Dyson's lab, one brushing mode, hard-to-reach areas only, and proxy plaque Dyson co-developed. 'Up to 69 percent' survives all four conditions and still gets to be the headline. The 470,000-image, six-year framing does similar work — averaged out it is about 78,000 images a year, a measure of effort standing in for a measure of accuracy. WIRED's closing shrug about whether the jets hit the right spots is the fairest gauge of the distance.
The seller owns every number
Dyson is the sole origin of the frame rate, the latency, the plaque advantage and the privacy assurance, and it has more riding on the launch than one hardware margin: proprietary paste, matched rinse, RFID heads that count their own expiry, and a subscription to deliver all three. Even the yardstick is partly in-house, built with the National University of Singapore's dental faculty. WIRED's own opening — Dyson enters a category and rivals must react — is the company's preferred myth about itself, reported accurately.
Sure what it is, unsure what it does
Two different confidences are tangled here. That the device exists, costs $499, reads 28 frames a second and needs foam-free paste is cleanly reported and internally consistent. Whether it beats a mid-price sonic brush in an actual mouth is untested by anyone in this reporting, as is the promise that no image is ever stored. Averaging solid description against unverified performance lands a little above the middle.