Cordless vacuum comparisons are decided by suction figures, and those figures are measured in ways that
have little to do with cleaning a floor. Air watts and pascals are recorded at the motor or at a sealed
inlet, not through a full-length wand and a floor head on carpet. The losses all happen in that last
stretch.
So this page averages five axes and weights none of them, which is a deliberately unflattering test. A
machine that is extraordinary at one thing and poor at another finishes below a machine that is merely
good at everything.
What an unweighted average actually rewards
Absence of a hole, and nothing else. The Shark tops this list without winning a single axis outright —
its performance, value and ease-of-use scores are identical and none of them is remarkable alone. The
V8 finishes second on flatness: a design from 2016 whose worst axis still sits at seven.
The two most capable machines here each arrive with one axis on the floor, and it is the same axis both
times. Miele builds the most durable body in the category and asks a premium for it. Samsung automates
the dirtiest step of the chore and asks a premium for that. Both give the money back on value, and both
land at the bottom of this five while cleaning better than the machines above them.
That is not a flaw in the arithmetic. It is the arithmetic doing what it says on the page.
Sealed filtration stopped being a flagship feature
Four of the five machines here route air through a path the maker describes as sealed. Dyson seals the
whole machine to a final HEPA stage. Miele runs a sealed airway into a lifetime HEPA filter. Samsung
stacks a multi-layer sealed system behind the dock. Shark rates its own as sealed HEPA and still trails
the other three on fine dust in reviewer consensus, which is a useful reminder that a sealing claim and
a fine-dust result are different objects.
The exception is the V8, whose two-tier radial cyclone predates the idea. On hard floors nobody notices.
In a house with an allergy it is the whole decision — and it is no longer a decision that forces you to
the top of the range.
Every runtime number on a box is for the lowest power mode with a non-powered tool. Samsung quotes a
hundred minutes, Miele quotes sixty, and neither number survives contact with carpet at full power,
where the honest figure across this category is under ten minutes.
No scandal in that, just physics. A motor pulling enough current to lift embedded grit from pile drains
a pack quickly.
What actually separates these machines is what happens when the pack runs out. Miele, Samsung and Shark
all list a removable battery, so a second one turns a whole-house clean into one session and a fading
cell into a purchase. The V8’s is fixed, which makes the same event a repair.
Weight is measured in the wrong place too
Stick vacuums are quoted at total weight, and distribution matters more than the number.
Miele is the clearest demonstration on this page. The block holding the motor and battery detaches and
remounts either high on the wand or low beside the floor head, and reviewer consensus is that the low
position changes how the machine feels far more than its eight-pound figure suggests. Put the mass below
your hand and pushing through carpet stops loading the wrist. Put it back on top and curtain rails get
easy again.
Every other machine here makes that choice for you and keeps it. Dyson mounts the weight high, which is
why the V15 is comfortable overhead and tiring on a floor, and why the lighter V8 is the one that gets
carried up stairs.
The dock moves a machine on two axes at once
Samsung’s station pulls the bin’s contents into a sealed bag, which removes the one moment in vacuuming
where the dust comes back at you. That earns the highest ease-of-use score on this page honestly.
It also has to live somewhere. The station and its cradle take permanent floor space beside an outlet,
reviewers flag the lead as short, and the bags are built for this station rather than sourced from
several suppliers, replaced on a schedule regardless of how full they look. Every one of those is an
ownership cost rather than a cleaning limitation — so the machine climbs one axis and drops another.
Averaging the two is the entire point of this ordering, and it is also why a different ordering exists.
The value ranking runs this same category on the value axis
alone. Only two of the five above survive it. The rest of that page is machines this one ranks below,
which is what happens the moment you stop averaging and read a single column.