How these comparisons are put together
Every comparison follows the same shape: the same ten attributes, read from the same registry, presented in the same order, followed by a verdict that names which unit suits most readers and which situation flips that answer. Nothing is weighted by a secret score.
PowerStationFacts takes part in the Amazon Services LLC Associates Program. When you buy through a link on this page we may earn a commission, and the price you pay does not change.
The ten attributes, always the same
Capacity, continuous AC output, surge rating, cell chemistry, cycle life with its endpoint, wall recharge time, maximum solar input, UPS switchover time, weight and battery expansion. These ten decide almost every real purchase, and holding the list fixed stops a comparison from quietly reshaping itself around whichever unit we would rather recommend.
Attributes outside that list appear in the written sections when they matter for the specific pair, such as an unusual outlet arrangement or a charging behaviour worth flagging.
How the score is built
Each unit carries a score out of 10. It is calculated, not awarded: six axes, each a quantity read from the registry, each normalised across this catalogue, then combined with a fixed weighting.
- Capacity, 25%. Usable watt hours as published.
- AC output, 20%. Continuous output, not surge. Surge figures are not comparable between makers, as the section below explains.
- Recharge speed, 15%. Watt hours per minute from a wall socket, so a large slow unit does not beat a small fast one on size alone.
- Cycle life, 15%. Rated cycles, adjusted for the endpoint. A rating quoted to 70% of original capacity promises less than the same number quoted to 80%, and is scaled accordingly before anything is compared.
- Watt hours per kilo, 15%. What you carry per unit of energy. This is what stops the scale from simply rewarding the heaviest unit.
- Backup switchover, 10%. Scored against practical thresholds rather than normalised: anything under 20 milliseconds will carry a desktop computer, and being faster than that wins nothing extra.
Three consequences worth stating plainly. The scale is relative to this catalogue: a 7 means seventh-tenths of the way between the weakest and strongest unit covered here, not an absolute verdict. A unit whose maker does not publish a figure loses that axis rather than scoring zero, and its page says what share of the weighting the score actually rests on. And a single number cannot express fitness for a purpose, which is why every unit also carries its rank within its own capacity class: a small unit scoring 5 may be the best small unit there is.
No manufacturer can influence a score, because nothing in the calculation is a judgement. Change a published specification and the score changes with it.
How surge ratings are handled
Manufacturers describe peak power in different ways, and the differences are not cosmetic. BLUETTI's Power Lifting and EcoFlow's X-Boost both run appliances above the unit's rated output by reducing voltage, which suits resistive loads such as heaters and kettles but not motors or electronics. A conventional surge rating means something else again: a brief tolerance for start-up current.
This site keeps the manufacturer's own label attached to the figure, so a Power Lifting number is never silently compared against a plain surge number as if they were the same measurement.
What would change a verdict
Verdicts follow from the registry. When a manufacturer revises a specification, the registry changes and every page that reads it changes with it. When a new model lands in a category we cover, it joins the comparison rather than being appended as a footnote.
If you find a figure here that disagrees with an official product page, that is a fault worth reporting, and the contact page is the place to do it.