What to Look For in a Grow Light

Run one check and make it this one: the fixture has to publish a photon figure in µmol/s. Everything else on a grow light box is either derived from that number or decorative. Of the eighteen fixtures this site tracks, the Spider Farmer SF1000 is the one we would buy on the strength of its specification sheet alone — it is the only fixture whose two headline figures divide into each other correctly — and it is the right answer for most people lighting an area of about two feet square.

We have not tested any of these fixtures. What follows is a procedure for reading what the manufacturer published, the arithmetic that turns it into an answer, and a scorecard showing how far each of the eighteen gets.

How we judged, before anything was ranked

Six checks, in this order. The order matters: each one is cheap to run and eliminates fixtures that the next check cannot rescue. Nothing here requires a light meter, a review or an opinion — only the specification page and a calculator.

1. A photon output figure in µmol/s (PPF).
2. An efficacy figure in µmol/J, published or derivable.
3. A PPFD figure with the measurement distance attached.
4. A coverage claim with a hanging height attached.
5. Dimming.
6. An honest account of what was left out.

Check 1: a photon figure in µmol/s

Plants respond to the number of photons that land on them, counted in micromoles per second. That figure — PPF, photosynthetic photon flux — is the only one that lets you work out whether a fixture can light your area before you own it.

Three figures routinely stand in for it and none of them substitute. Lumens weight light by human eye sensitivity, which peaks in the green a plant reflects. Watts measure what goes in, not what comes out — the whole point of the next check is the ratio between them. “Equivalent watts” is marketing against a lamp nobody sells any more; we take that one apart in watts vs equivalent watts.

If it is missing: you cannot size the fixture, and no amount of reading will fix that. Ten of our eighteen publish a PPF. Barrina, whose strips are the default seed-starting bar in this category, publishes none for any product in its range, which is the reason this site has a coverage calculator at all.

Check 2: efficacy in µmol/J, and the bar it should clear

Efficacy is photons out divided by joules in — PPF divided by watts. It decides what the fixture costs to run for a given amount of light, and it is the single number that separates a 2026 LED from a 2016 one.

There is an independent benchmark for it, which is the useful part. The DesignLights Consortium maintains a qualified products list for horticultural lighting, and its Version 4.0 requirements, effective 18 April 2025, set a minimum photosynthetic photon efficacy of 2.5 µmol/J— an 8.7% increase on the previous version, measured to ANSI/IES LM-79-19 at an accredited laboratory.

Three of our eighteen clear it: the AC Infinity IONBOARD S22 at a published 2.7 µmol/J, the Spider Farmer SF1000 at exactly 2.5, and the VIVOSUN VS1000 at 2.75 — and the VS1000 no longer appears in Amazon’s catalog, so in practice it is two. That is not a scandal: the DLC bar is written for warehouse-scale luminaires, and a screw-in bulb pays a real penalty for being a bulb. But it does tell you the size of the penalty, which no listing will.

Which is the honest frame: the penalty is real and it is small in absolute terms. We rank by it in the µmol/J table, and it is not the reason to reject a bulb. Lighting one plant with a panel is.

Check 3: a PPFD figure with the distance attached

PPFD is what arrives at the leaf, and it falls off roughly with the square of distance. A PPFD number without the distance it was measured at therefore carries no information at all — it is true at some height, and the manufacturer has not said which.

This is the most common tell in the category. AC Infinity publishes “1098 PPFD” for the IONBOARD S22 with no measurement distance anywhere on the page. We cannot use that figure and we do not try to. Three fixtures of our eighteen do it properly: SANSI at 265.58 µmol/m²/s at 1 ft, Soltech at three separate distances, and the VIVOSUN LumaLight at 521 µmol/m²/s at 12 in.

If it has a distance, you can move it. That is the whole value of the check.

Check 4: a coverage claim with a height attached

“Covers 2 × 2 ft” is not a specification until you know from how far away, and at what intensity. Mars Hydro does this well by accident of honesty — it publishes 2.5 × 2.5 ft for vegetative growth and 2.3 × 2.3 ft for flowering, which at least admits that coverage depends on how much light you need. ViparSpectra goes further and attaches the heights: 2.5 ft square at 16 inches, 2 ft square at 10 inches.

A coverage figure with no height is a statement about where the light stops being visible, not where it stops being useful. If that is all you have, ignore it and do the division yourself — what size grow light do I need is that division, worked.

Check 5: dimming

Dimming is the one feature on this list we would pay for, and the reason is arithmetic rather than convenience. Every sizing decision on this site has error bars: published figures carry no tolerance, your area is not the manufacturer’s test area, and plants move. A dimmer is the correction, and it is free after purchase.

It also changes the failure mode. Buy a fixture too small and you are stuck; buy one too large with a dimmer and you turn it down. Eight of our eighteen dim, and every LED board except none of them does — all five boards have a knob or stepped control. No grow bulb we cover is dimmable, and putting one on a wall dimmer is how you get buzz and flicker.

Check 6: what they did not publish

The last check is a reading habit rather than a number. Write down what is absent. A manufacturer that publishes a PPF, an efficacy, a PPFD with a distance and a rated life has told you four things that can be checked against each other; one that publishes a coverage area and a lumen count has told you nothing that can be checked at all.

Our favorite instance: Spider Farmer publishes 249.21 µmol/s and 2.5 µmol/J for a 100 W fixture. 249.21 ÷ 100 = 2.4921, which rounds to the figure they printed. The two numbers were measured, not chosen. That consistency check takes four seconds and almost nothing in this category survives it.

The scorecard, all eighteen fixtures

Every cell below is from the manufacturer’s own specification page. “Ours” marks an efficacy we divided out of two figures they did publish. Blank means they published neither.

Ten of eighteen publish a photon figure. Twelve have a knowable efficacy. Three clear 2.5 µmol/J and one of those three is out of stock everywhere. Three publish a PPFD you can move. That is the state of consumer grow light specification in 2026, and it is why a procedure beats a shortlist.

1. Spider Farmer SF1000 — two numbers that divide into each other

Clears checks 1, 2, 4 and 5, and is the only fixture here whose published figures validate one another. 249.21 µmol/s from 100 W, published alongside an efficacy of 2.5 µmol/J, which is what the division gives. It meets the DLC minimum exactly.

Best for: anyone lighting roughly two feet square who wants the specification to be checkable rather than impressive.

The downside: no PPFD at any distance, so you cannot predict what arrives at the leaf without doing the division from PPF and your own area. Spider Farmer also quotes a 2 × 2 ft core and a 3 × 3 ft maximum with no height for either, so check 4 is passed on a technicality.

2. Mars Hydro TS 1000 — the largest published photon figure

343 µmol/s is the most light anyone here publishes, and 150 W is the most electricity. At 2.3 µmol/J it sits below the trade bar, which is the trade-off stated plainly: more photons, slightly worse photons per watt.

Best for: the largest area a single consumer fixture can reasonably cover — a 2.5 ft square, or a wide shelf.

The downside: it misses 2.5 µmol/J, and 150 W run 16 hours a day is 72 kWh a month, about $13.20. It is the most expensive fixture here to run, and the coverage figures split by growth stage in a way that flatters the larger number.

3. AC Infinity IONBOARD S22 — the highest efficacy anyone publishes

2.7 µmol/J, published, clears the DLC minimum with room to spare, and dims in 20% steps. From the efficacy and the wattage you can back out roughly 311 µmol/s, which we label as ours because AC Infinity did not print it.

Best for: long runs — a fixture that will be on twelve hours a day from October to March, where the efficiency compounds.

The downside: it fails checks 1 and 3. No PPF, and a “1098 PPFD” claim with no measurement distance attached, which is unusable. The best-specified fixture here by efficacy is one of the worst-specified by everything else.

4. SANSI 36W BR30 — the cheapest fixture you can place by arithmetic

The only bulb in the category that publishes both a PPF and a PPFD with the distance attached, which is why it appears on more pages of this site than anything else. 65.6 µmol/s, 265.58 µmol/m²/s at 1 ft, 1.82 µmol/J once you divide.

Best for: one plant, in a lamp you already own, where you want to know in advance what height to set.

The downside: 1.82 µmol/J is well under the trade bar, no beam angle is published so the 265.58 figure cannot be turned into a footprint, and SANSI states the bulb weighs a pound and is “not recommended to be used in tightly closed fixtures”. Which lamp it goes in is a real constraint — the fitment checks are here.

5. GE Grow BR30 — a photon figure under $25

Two bulbs, 16 µmol/s each, 9 W each, E26, and a mainstream brand that printed the photon number on the box. That is the whole case, and for the price it is a good one: the cheapest way to own a fixture whose output is a known quantity rather than a claim.

Best for: two ordinary lamps, two plants, and a buyer who wants arithmetic rather than a panel.

The downside: 16 µmol/s is a very small amount of light. Even if every photon landed inside one square foot — which no bulb manages — the average would be 172 µmol/m²/s, and GE publishes no beam angle to check that against. Non-dimmable, and no enclosed-fixture statement anywhere on the page.

Three claims to ignore outright

• “Full spectrum.” No threshold exists for the phrase, so printing it costs a manufacturer nothing. The long version is in what full spectrum actually means.
• Lumens, and anything in lux. Human-eye units. There is no universal conversion to photon units because the ratio depends on the spectrum of the specific fixture.
• “Replaces a 600 W HPS.” A comparison against a lamp with a different spectrum, a different fixture geometry and a thirty-year-old efficiency. Nothing is being measured.

What this procedure does not tell you

Three things, and they are worth naming because the checks above can make a specification sheet feel more complete than it is.

Uniformity. A PPF figure says how many photons leave; it says nothing about whether they land evenly. Two fixtures with identical PPF can produce a bright center and dark corners or an even field, and almost nobody publishes a map. The practical consequence is in bar vs panel coverage.

Whether the numbers are true. We check published figures for internal consistency, which catches carelessness and invention. It does not catch a manufacturer who measured honestly on a unit better than the one in the box. We have not tested these fixtures and we say so on every page.

Longevity. Rated life figures here run from 25,000 to 50,000 hours, and one fixture quotes 36,000 hours to Q90 — the point at which output has dropped a tenth. Those are projections from standardised tests, not observations of the fixture you will own.

The verdict, compressed

If the listing has a PPF in µmol/s and a wattage, you can decide everything else yourself. If it has an efficacy above 2.5 µmol/J you are buying 2026 hardware. If it has a PPFD with a distance you can place it without a meter. If it dims, you can be wrong about your area and recover.

For most people reading this, the SF1000 clears enough of that to be the default, the SANSI bulb is the right answer for one plant, and the fixture to walk away from is the one whose page is a list of adjectives.