Somewhere on this page there is a thirty-dollar light claiming ten thousand lumens, sitting next to a hundred-and-twenty-dollar light claiming two thousand eight hundred. One of those numbers was measured against a published standard and one of them was not, and if you cannot tell which, the flashlight market will take your money for the next decade.
So this page leads with the measurement problem, because the measurement problem is the buying advice. Once you know what to compare, the ranking mostly writes itself.
Three numbers, and only one of them is on the box. FL 1 lumens are measured at the front of the lens thirty seconds after switch-on, to the ANSI/NEMA standard. A marketing lumen number is usually the emitter's theoretical maximum at the instant of switch-on, which nobody ever sees. Sustained output is what the light holds after it heats up — every small high-output light steps down, and the honest makers publish the curve. Candela is beam intensity, and it is what decides how far the light reaches: a 2,000 lumen flood can be less useful at distance than a 600 lumen thrower. Everything on these pages is ranked on the second and third of those, never the first.
Why lumen inflation is worse here than anywhere else
Two things make flashlights uniquely bad for this. The first is that lumens are invisible: you cannot look at a beam and know whether it is 800 or 1,600, and human vision is logarithmic, so doubling the output looks like a modest increase rather than twice as much. A buyer has no way to audit the claim by using the product, which is not true of a knife or a tent.
The second is that there is a technically-true-ish number available to quote. An LED emitter has a rated maximum output at a given current, and a manufacturer can print that figure without ever measuring the assembled light. It ignores the reflector, the lens, the driver and the thermal design — all of which take real light away. The gap between emitter rating and FL 1 measured output is routinely a factor of two or more.
The ANSI/NEMA FL 1 standard exists precisely to close that gap. It specifies how output, beam distance, runtime and impact resistance are measured. Manufacturers who follow it can be compared to each other. Manufacturers who do not, cannot be compared to anything, and that absence is the most useful specification on the listing.
The thermal step-down nobody advertises
Every light on this page will produce its headline output. Almost none of them will produce it for more than a couple of minutes.
The reason is physics rather than deception. An LED at high current is mostly a heater, and a light small enough to carry has nowhere to put that heat. So the driver steps the output down to protect the emitter — typically starting somewhere between thirty seconds and three minutes in, and settling at a fraction of the peak. That fraction is the number that matters, because it is the light you actually have while you are using it.
The tell is simple: does the manufacturer publish a runtime chart? Fenix, Streamlight and Nitecore do. A chart with a visible cliff in the first two minutes is not a confession, it is an honest description of a real light. A single number with no chart is the thing to be suspicious of.
Candela, and why a dimmer light can reach further
Lumens measure total light output in every direction the light throws it. Candela measures intensity in the middle of the beam. They are different questions and they have different answers.
A wide flood beam spreads its lumens over a large area, which is what you want for working on something in front of you. A tight beam concentrates them, which is what you want for identifying something a hundred yards away. This is why Streamlight and SureFire publish candela alongside lumens and why the enthusiast brands increasingly do too — without it you genuinely cannot predict how a light will behave.
The practical rule: for close work, buy lumens. For distance, buy candela. If a listing gives you only lumens and you need distance, you are guessing.
These are researched picks, not lights I carry. Elsewhere on this site there are pages about gear I actually own, and those are written in the first person because they have earned it — my rifles, my truck, my computer. Flashlights are not one of those categories yet. So this cluster does what the rest of the site does when I have not handled something: reads the published specifications, compares the numbers that are actually comparable, says which manufacturers publish enough to be checked, and names one to skip.
Where this fits with the rest of a pocket
A light is the third item in the pocket trio — the other two being a knife and a wallet, and both of those have their own pages here. If you are building the whole loadout rather than replacing one item, the complete minimalist EDC page puts them together, and best EDC flashlights covers the keychain and pocket end of the category in more detail than this page does.
How to buy without being lied to
Look for a runtime chart before you look at anything else
If the manufacturer publishes one, you can see the step-down and compare honestly. If they do not, you are being asked to trust a single number with no method behind it. This one test eliminates most of the category in about ten seconds.
Compare the sustained figure, not the turbo figure
Turbo is a marketing mode with a legitimate use — a short burst to identify something. The number you will live with is the high or medium setting the light holds indefinitely. Ranking on turbo is how people end up disappointed by a light that met its specification.
Decide whether you need throw before you shop
Close work wants flood; distance wants candela. Most people buying a "tactical" light are actually buying a general-purpose light and would be better served by a balanced beam than by either extreme.
Prefer standard cells and standard charging where you can
Onboard USB-C is convenient. Proprietary magnetic docks are more convenient still, right up until you are somewhere without the dock. A light that takes a common cell and charges over a common cable is one you can keep running anywhere — see the battery platform page for what that actually means.
Buy a cheap light from a brand that publishes numbers
The alternative to an expensive light is not an unbranded listing with a fictional lumen claim. It is Wurkkos or Sofirn, who publish real specifications at real prices. There is no longer any reason to buy fiction to save money.
