Power station specifications, compared against the class
A power station prints two numbers on the box and explains neither. Watts decide whether an appliance runs at all; watt-hours decide how long. This site computes what they mean together.
Most power station advice tells you which one is best. This site tells you what the market looks like, so you can see where any particular unit sits in it — and what its two headline numbers actually imply. The median station here holds 1,024 Wh and runs 51 min at its own rated output.
Start with the finder
Four questions about what you need to run and for how long, and the power station finder narrows the catalogue. Every result carries the runtime its own specifications imply.
The finding this site exists to publish
Across every output class in this corpus, the median station runs about fifty minutes at its own rated output. Not the small ones for less and the large ones for more — about the same, because capacity and output rise together as you go up the range.
A bigger station does not buy you a longer run at full tilt. It buys you the ability to run more things, or bigger things, for about the same hour. That is genuinely counterintuitive, it falls straight out of dividing one published number by another, and no listing does the division for you.
The runtime figure is arithmetic, and it is optimistic. It is the published capacity divided by the published output, which is the best you could ever get. Real inverters are 85 to 90% efficient, so the usable figure is lower — and no listing on this site states its inverter efficiency, so we do not estimate it. Treat the number as a ceiling and a way to compare like with like, not as a promise.
What a baseline is, and why it is the whole point
A single average across every power station would be meaningless. This corpus runs from pocket batteries to home backup systems, and someone shopping a camping station is not served by a number that includes both. So baselines are computed per output class, per capacity band and per cell chemistry — each axis on its own, because 4 classes by five bands is twenty cells over 170 products and almost none would clear the floor.
That produces 11 baselines across 11 groups. Where a group holds fewer than 8 listings, no baseline is published at all.
What we do not do
We do not test power stations, and we say so on every page. There is no lab here, no load bank, no meter. What we have is the whole catalogue read consistently, which is a different kind of useful and an honest one.
We do not publish star ratings or review counts, and we do not print exact prices. Where a listing does not state a specification, we print not stated rather than a plausible default.
How the corpus is cleaned
Search results for power stations contain uninterruptible supplies, jump starters, bare solar panels and carry cases. The hard case is the UPS, because many genuine power stations advertise a UPS mode — so the test is what the listing calls itself, not what it mentions. The methodology page documents exactly how, including where the rule fails.
About this site
Power Station Data is written by a named editor working from published specifications, and funded by affiliate commission on Amazon purchases. Both facts are documented rather than buried:
- About Power Station Data — who writes this, and what a pen name does and does not mean.
- How we compile these figures — the full method, including where it fails.
- Affiliate disclosure — how this site earns, and what that does not influence.
- Contact — corrections to published figures are the most useful thing you can send.
- Privacy policy — what is collected, which is very little.