Sofirn SC31 Pro Review: Two Thousand Lumens, No Runtime Printed

The verdict in sixty words

This light sits at the top of our flagship ranking, our under-fifty ranking and the Ledger, which is exactly why it has earned a page that argues with it rather than one that agrees with it. Everything below is read from Sofirn's own published specification, with a single independent measurement clearly marked as somebody else's work rather than ours.

What Sofirn publishes

Sofirn sells direct and has no retail shelf to pay for, and that shows up in the specification more than it does in the price. For roughly what a supermarket charges for a sealed plastic torch, the published sheet reads: 2,000 lumens maximum from a single SST40 emitter, 226 meters of beam distance, IPX8 water resistance, 59 grams without the cell, 117.5 millimeters long, USB-C charging that Sofirn states takes two and a half hours with a 2A adapter, and one 18650 lithium-ion cell the owner can change.

The row that is blank

Two rows on that table are empty, and on this site those matter more than any of the rows that are full. Sofirn does not publish how long the SC31 Pro runs at 2,000 lumens, or at any other output, and it does not state that its figures were measured to ANSI FL1. Every ranking here is arithmetic over published numbers, so a missing number is not a gap in our research. It is a fact about the product.

This is not pedantry, and here is why. An aluminum tube of this size cannot shed the heat that 2,000 lumens produces for very long. Nothing in this class can, and the physics does not care which brand is on the label, so every light like this steps down. The only question is whether the manufacturer tells you when. Wurkkos prints one hour and fourteen minutes beside the TS10's 1,400 lumens. NEBO prints thirty seconds beside the SLIM Mini 500's turbo. Sofirn prints nothing.

The practical reading is to treat 2,000 lumens as a burst rather than a mode, and to buy the light for the sustained output underneath it. That is how it will behave in your hand and it is also how it is priced, so this is a criticism of the specification sheet rather than of the flashlight.

The independent measurement, and what it does not settle

There is one piece of outside evidence worth putting on this page. A reviewer on BudgetLightForum posting as gchart measured this light on his own equipment and reported 2,434 lumens at turn-on and 2,255 lumens at thirty seconds. If that holds generally, Sofirn's 2,000-lumen claim is conservative, which is close to unheard of in a category where the number on the box is usually the number the emitter can touch for an instant.

We are not going to overstate it. That is one reviewer, one sample and one set of instruments — not a certified test and not a batch. It says nothing about output at ten minutes, which is the figure that would actually settle the runtime question. And the same reviewer found something less flattering: the thermal calibration on his unit was off by roughly 22 degrees Celsius from the factory, which makes the light step down earlier than it should until the owner corrects it in the firmware. A light whose thermal behavior depends on the buyer knowing about a configuration menu is a light that ships unfinished for most buyers.

The cell is the actual product

Strip the marketing away and what you are buying for around thirty dollars is a well-made aluminum tube built around a commodity 18650. That sounds like a criticism. It is the opposite, and it is the single reason this light leads our rankings.

An 18650 is an industry-standard cell made by dozens of manufacturers and sold everywhere. When the one in this light has given up its capacity — three years, five years, depending on how hard you cycle it — you unscrew the tailcap, drop in a new one, and the light is effectively new. Compare that with a sealed light at the same price: when its battery reaches the end of its cycle life, the emitter, the driver, the optic, the switch and the machined body all go in the bin with it. Two lights with identical lumens-per-dollar figures on the day of purchase can differ threefold over a decade, and this is where that difference comes from.

IPX8 is unusual at this price

IPX8 means the manufacturer states the light survives continuous immersion beyond one meter, at a depth and duration the manufacturer itself defines. Most lights at this price publish IPX4, which is splash resistance, or publish nothing at all — and a missing water rating is its own kind of answer. Sofirn publishing IPX8 here, against Wurkkos publishing IPX7 on the FC11, is one of the few places where the SC31 Pro wins on documentation rather than on output. Our guide to IP and IPX ratings sets out what those digits do and do not promise.

Anduril 2 is the price of admission

The SC31 Pro ships with Anduril 2, the open-source firmware the budget enthusiast brands have standardized on. It gives you smooth brightness ramping instead of four fixed steps, configurable shortcuts, a battery-check readout, a lockout, and a deep configuration menu that includes the thermal calibration mentioned above.

It is also, for the first week, irritating. Click counts do different things depending on which state you start from, and nothing in the box makes that obvious. People who like flashlights find this delightful; people who simply want a flashlight find it exhausting. That divide is real and we are not going to pretend it is a footnote — it is the most common reason somebody sends this light back, and it has nothing to do with how well the light works.

When to buy something else

Three specific cases where this is the wrong purchase, in the order they come up most often. None of them is a knock on the light; they are all a mismatch between what it is and what the buyer wanted.

• You want one button and one brightness. Buy a simple three-mode light and enjoy it. Anduril will not grow on you if you have already decided it will not.
• You need colors to read correctly — wiring, paint, an engine bay, a face. The SST40 is an output emitter, not a color-accurate one. The Wurkkos FC11 gives up seven hundred lumens for a 90-CRI emitter and is the better buy for close inspection work.
• You want it in a front pocket all day. At 117.5 millimeters this is a jacket-pocket or bag light. Sofirn's own SC21 Pro is the smaller answer, and our EDC rankings cover the rest of that shape.

How we scored it

Four attributes, each out of ten, weighted as published on our methodology page: output against price, battery architecture, build as evidenced by the published ratings, and usability. The SC31 Pro scores at or near the top of the first three of anything on this site and is marked down on the fourth, because Anduril costs a real number of buyers a real amount of patience. One further adjustment applies here, as it does anywhere: a light whose headline output carries no published runtime is scored below one whose does. It still leads, which is a measure of how far ahead the other three attributes are rather than a sign we let the missing row slide.

What we cannot tell you

We have not had this light in our hands. We cannot tell you what the tint looks like beside a warmer emitter, how the knurling feels after a winter, whether the anodizing survives a toolbox, whether the threads are smooth out of the box, how consistent one unit is against another from a different batch, or what Sofirn's support is like when something fails. Budget brands are most variable on precisely those points. Where somebody measures properly — the BudgetLightForum thread cited below runs real instruments — read them alongside this page and treat our contribution as reading the specification carefully rather than testing the product.

The bottom line

If you have decided to spend around thirty dollars on a rechargeable flashlight and you are willing to read a firmware manual once, this is the light with the strongest published case in the category: the most lumens per dollar we rank, the strongest water rating in its price class, and a cell you can keep replacing indefinitely. If you would rather press a button and get light, or you need colors to read true, buy one of the alternatives above instead. The ranking is a piece of arithmetic, not an instruction.