The box prints a lumen number. It almost never says when that number was read, how fast it falls, or what color the light is. Here is the standard behind the numbers, the battery formats, the nine lights I own, and five I don't that independent meters have checked.
The rulebook is ANSI/PLATO FL 1, and Streamlight's summary gives the three definitions that matter. Lumens are measured 30 seconds after switch-on with fresh batteries. Beam distance is where the beam falls to 0.25 lux. Run time counts from that 30-second value until output drops to 10 percent.
So a light has three lumen numbers: turn-on (the box), 30 seconds (the standard), and sustained (what you carry once heat forces a step-down). The runtime rule is the loophole. A light that drops from 1,500 to 470 inside a minute can book hours at 470 as turbo runtime. 1Lumen's X4 test shows it: 1,460 at turn-on, 1,053 at 30 seconds, 469 at ten minutes, then an ANSI turbo runtime of 3 h 30 min spent almost entirely near 470. The rule is fine with that. Throw is derived, not paced out: 1Lumen's candela explainer gives it as the square root of candela over 0.25 lux, so 7,950 measured candela is 178 meters against 205 claimed.
Who follows the standard? 1Lumen's lumen guide says only a handful of makers. PLATO's FL 1-2025 revision, published February 2026, adds a separate method for short turbo bursts. The boxes have not caught up. Until they do, the check is the meter crowd. 1Lumen reads lumens at turn-on, 30 seconds, and ten minutes, in a sphere or a home-made lumen tube depending on the reviewer; its X4 test used a tube. ZeroAir calls his gear hobbyist-level and a 10 to 15 percent gap normal. BudgetLightForum members post Opple readings. Who is in the meter crowd, and what each tester measures: the flashlight reviewers who measure.
| Light | Box claim | Meter at 30 s | Throw, claimed → measured | Who measured |
|---|---|---|---|---|
| Wuben X4 (owned) | 1,500 lm | 1,053 lm | 205 → 178 m | 1Lumen |
| Olight Baton 4 (owned) | 1,300 lm | 1,441 lm | 170 → 168 m | 1Lumen |
| Olight Oclip (owned) | 300 lm | 274 lm | 70 → 84 m | 1Lumen |
| Olight i3E EOS (owned) | 90 lm | 114 lm | 44 → 47 m | 1Lumen |
| Wuben E2 Pulse (keychain) | 150 lm | 84 lm | 26 → 30 m | ZeroAir |
| Nitecore TIP3 (keychain) | 720 lm | 627 lm | 92 → 89 m | ZeroAir |
| Wurkkos FC11C | 1,200 lm | 975 lm | 155 → 156 m | 1Lumen |
| Acebeam E75 (Nichia 519A) | 3,000 lm | 2,767 lm | 210 → 211 m | 1Lumen |
| Fenix PD36R Pro | 2,800 lm | 2,494 lm | 380 → 370 m | 1Lumen |
My compilation from the linked reviews. Different rigs, so judge a light against its own claim, not against its neighbors. The keychain lights miss by the most. Olight runs the other way: two of my three measured Olights beat the number on the box at 30 seconds, and the Oclip missed it by nine percent. That is not how this table usually goes.
CRI scores how faithfully a light shows color against a reference, 0 to 100, with incandescent near 100. Ra averages eight pastel test colors. Strong red (R9) is scored separately, and it is what LEDs get wrong first (Wikipedia's CRI entry covers the samples and the metric's limits).
The X4's main LEDs read 64 CRI on worrub918's Opple meter at BudgetLightForum (77 on the warm side light); 1Lumen's meter agreed at 64.8 and a cool 5,815 K. My Olights are no better on color: 1Lumen read the Baton 4 at 63.6 CRI and 5,583 K, the Oclip at 65.7, the i3E EOS at 69.1. So my four measured lights all land between 63.6 and 69.1 CRI. Consistent, at least. At that level, reds and browns flatten: fine for finding a dropped screw, poor for telling a brown wire from a red one or judging a leak. BudgetLightForum's own Opple thread calls it a consumer-grade sensor with approximate CRI, so read hobbyist figures as plus or minus a few points.
High CRI here almost always means a Nichia 519A. 1Lumen measured the Wurkkos FC11C at 96.8 CRI and the 519A Acebeam E75 at 96.5; ZeroAir read the high-CRI Pokelit AA at 90 or better on every mode. The trade is lumens: Acebeam rates the same E75 at 3,000 with 519As and 4,500 with cool-white LEDs. Tint (6,500 K bluish, 4,000 to 5,000 K neutral) is separate and buys no accuracy. The TIP3 offers three tints, and ZeroAir measured all of them near 70 CRI.
Replaceable 18650. The X4 and FC11C run a standard 18650, so the light outlives its cell and a spare rides along. Check button-top versus flat-top (the X4 takes either). 21700 is the bigger sibling, with more capacity and a higher discharge rate per 1Lumen's battery guide. That is why the E75 and PD36R Pro hold output the X4 cannot.
AA and 14500. The Pokelit AA takes a 14500 lithium cell or a plain AA: same light, half the output, 460 lumens against 179 in ZeroAir's test. Use NiMH. 1Lumen's guide refuses to recommend alkaline AAAs, and AA is no different.
Sealed cells. Most keychain lights run on one: 650 mAh in the TIP3, 120 mAh in the E2 Pulse, and ZeroAir's main complaint on both. My Oclip, iMini 2, Yemao and AOLBSPD are sealed too. When the cell fades, the light is done. The i3E EOS is the exception, on one AAA you can buy anywhere.
USB-C. Most lights here charge from any USB-C source at about 5 V and 1 to 2 A, so a phone brick is plenty (see chargers and power or the AOHI 140W I use). The Baton 4 wants Olight's magnetic cable instead, and the i3E wants a fresh AAA. The catch is time. 1Lumen clocked the X4 at 3 h 7 min. The listing says 1.5 hours. One of those was measured.
Keychain. About an ounce (the TIP3 is 1.06 oz), usually a sealed cell, claims of 100 to 700 lumens. The lock, the glovebox, the screw under the desk. Pairs with the BIBURY keychain multitool on the EDC gifts list.
Pocket EDC. An 18650 or 21700 in a tube or flat body, 1,000 to 3,000 lumen turbo claims, 300 to 1,000 sustained, a clip, often a tail magnet. The X4, FC11C, and E75: the light next to a pocket knife.
Tactical. Forward-clicky tail switch, momentary on, instant strobe, throw-first optics. The PD36R Pro's 34,200 candela (370 m) against the X4's 7,950 (178 m) is the whole difference. My JECKLAMP is tactical by listing title only.
The full write-up on the light behind this page's numbers, including the listing that calls its battery sealed. It isn't.
USB-C bricks and power banks that keep an 18650 light topped up.
Nine lights. Four are Olights, and three of those have been through 1Lumen’s meter. The Wuben X4 has the fullest data of any light I own. The other four are Amazon-only brands: big claims, no meter, and a listing to go on.
Flat-body 18650, USB-C, rotary dimmer, RGB side light. 1Lumen: 1,053 lm at 30 s, about 470 sustained, 178 m, 64.8 CRI. The 470 you get to keep. The CRI you cannot fix. Full write-up in the X4 review.
For: one pocket light for throw, flood, and camp duty with a swappable cell
Skip if: you need accurate color or sustained output
Check price on AmazonA pocket tube on a 16340 cell with Olight’s magnetic MCC charging, five levels from a half-lumen moonlight to a claimed 1,300. 1Lumen measured 1,486 lm at turn-on, 1,441 at 30 s and 303 at ten minutes. Throw came in at 168 m against 170 claimed, color at 63.6 CRI and 5,583 K, weight at 54 g with the cell. The rare light that beats its own box. Mine is the orange.
For: the smallest light here that still throws; clip and tail magnet
Skip if: you want a standard cell (the 16340 is Olight’s own) or high CRI (1Lumen read 63.6)
Check price on AmazonA clip light: white and red LEDs in a 28-gram clip that grips anything up to 14 mm, with a magnet in the tail. Claimed 300 lumens; 1Lumen measured 279 at turn-on and 274 at 30 s, with 84 m of throw against 70 claimed and 65.7 CRI. It ran 1 h 02 min to ten percent on high, and the sealed 280 mAh cell took a 93-minute USB-C charge.
For: a hat brim, a strap, the edge of a page: light where a tube cannot clip
Skip if: you need more than an hour on high, or a cell you can replace
Check price on AmazonOne AAA, a twist head, one mode, 19 grams. Claimed 90 lumens and 44 m. 1Lumen measured 114 lm at 30 s and 47 m on the included alkaline, 116 lm and 48 m on NiMH, 69.1 CRI. Runtime to ten percent: 49 minutes on alkaline, 1 h 20 on NiMH. A keychain light whose cell you can buy at a gas station. Mine is the black.
For: keys, and anyone who distrusts sealed cells
Skip if: you want modes, or a clip
Check price on AmazonA magnetic keychain light with a built-in 10180 cell and a USB plug under the cap: pull the body off the cap and it is on. Olight claims 50 lumens, 21 m, 60 minutes of runtime and a 60-minute charge, 0.69 oz. No meter has checked this one, so those stay claims. Mine is the OD green.
For: the smallest always-with-you light
Skip if: you want a number a meter has read; the i3E above has one
Check price on AmazonListing claims: 600 lumens, six modes, built-in rechargeable cell, about 1.65 inches, steel chain. The brand has a BudgetLightForum footprint (a member review of its A200), but no meter has checked this model. Until one does, 600 is a number on a listing.
For: the keyring, where a sealed cell is the norm
Skip if: you want a verified claim: buy the TIP3 below
Check price on AmazonAmazon-only brands with no presence at 1Lumen, ZeroAir, or BudgetLightForum, and claims of the kind that missed by half in the table above. I use them, and they light things up. How many lumens that is, no meter has said, and I am not taking the listing's word for it. No link until one does.
| Light | What the listing claims | Why it waits |
|---|---|---|
| AOLBSPD 3-in-1 keychain flashlight | 500/1,200 lumens, red-blue strobe, phone recharge over USB-C, bottle opener, package cutter, IPX7 | 1,200 lumens from a keychain-sized sealed cell, unread by any meter |
| Jawploy pocket flashlight | 675 lumens, dual top and side light, seven modes, 36 g, magnet and clip, USB-C, IPX4 | Flat dual-source design in the X4's mold; a brand with no outside test anywhere |
| JECKLAMP tactical flashlight | 1,270 lumens, six modes, USB-C cable included, IPX8 | Tactical by title; no tail-switch spec, no throw figure, no outside test |
These come from the reviewers above, not my pocket. I don't own any of them, so there is no ownership label and the numbers are theirs. Each Amazon link was checked against the exact model the tester measured.
ZeroAir: 627 lm at 30 s against 720 claimed, 89 m, three tints, 650 mAh sealed cell, USB-C, 1.06 oz. His knocks: about 70 CRI and the sealed battery.
For: a keychain light whose claim holds up
Skip if: you want high CRI or a replaceable cell
Check price on AmazonThe Nichia 219F version. ZeroAir: 460 lm on the included USB-C 14500, 179 on an AA, 97 m, 90-plus CRI on every mode. Not the 1,000-lumen cool-white Pokelit, which 1Lumen measured at 70 CRI.
For: a pen-sized light that runs on gas-station AAs in a pinch
Skip if: you want turbo; this is a 460-lumen light
Check price on AmazonNichia 519A at 5,000 K on a regulated buck driver. 1Lumen: 975 lm at 30 s, 156 m, 96.8 CRI, five stars; BudgetLightForum's SammysHP logged flat output for most of the runtime. Swappable 18650, USB-C, tail magnet.
For: the high-CRI answer to the X4 at the same cell size
Skip if: you want a flat body or a side light
Check price on AmazonFour 519As on a 5,000 mAh 21700. 1Lumen: 2,767 lm at 30 s against 3,000 claimed, 211 m, 96.5 CRI, over 1,000 lumens held three hours on High; ZeroAir read about 95 CRI and flat output.
For: sustained high-CRI output, not a one-minute turbo
Skip if: 7.3 oz is too much pocket
Check price on AmazonLuminus SFT70, 21700, dual tail switches. 1Lumen: 2,494 lm at 30 s against 2,800 claimed, 34,200 cd for 370 m, 3 h 21 min ANSI turbo runtime, five stars. A 70 CRI work light by design.
For: throw, momentary on, and strobe under one thumb
Skip if: color rendering matters; buy the E75
Check price on AmazonOne email when a new guide or measured test lands. No spam, never sold, unsubscribe anytime.