Identify a gemstone by color, then know the limit

A gem identifier reads color, transparency, luster and cut, and returns the likely species. That is genuinely useful and genuinely limited: no photo and no home check sees heat treatment, lab origin or value. The key below gives you the species shortlist and says plainly where a stone earns a lab.

Which gemstone is this: work the key

Start with color, because that is what a cut stone shows first, then narrow by transparency and by what a ten-power loupe finds inside. Every check this key asks for is non-destructive: no scratch tests, no heat, nothing you would regret on a set stone or on something that is not yours. Most branches end on a group of two to four species rather than a single name, and that is the honest answer — a blue faceted stone genuinely is compatible with sapphire, spinel and topaz at once.

Step 1
In daylight, against white paper, what color is the stone?
Never scratch-test a set stone, and never test one you do not own. Everything this key asks for is done with your eyes, a loupe and a UV lamp.
Step 2
Red and pink: how much light passes through?
Step 3
Blue and violet: does the color hold under a warm lamp?
Step 4
Green: what does a loupe show inside?
Step 5
Yellow, orange and brown: how does it feel in the hand?
Step 6
Colorless: put a loupe on it and look through the table at the back facets.
Step 7
Black and very dark: what is the weight and the break like?
Step 8
Optical effects: what exactly moves when you tilt the stone?

Start over

A loupe does most of the work here and costs less than a takeout meal. The two observations worth learning are bubbles, which retire glass instantly, and doubling of the back facet edges, which separates zircon, peridot and moissanite from diamond, spinel and garnet.

Candidates, home checks and lab questions

The third column is the one no competing identifier prints. Every row states which question about that stone a laboratory has to answer: whether the color came from heat, whether the crystal grew in the ground or in a factory, whether the emerald is filled. Those questions carry most of the price of a gem, and no photograph and no kitchen test has ever answered one of them.

Candidates a photo allows	What you can check yourself	What only a lab settles
Ruby, red spinel or red garnet	A UV lamp separates most of the group: ruby usually glows red, while iron-bearing red garnet stays dark. Ruby also tends to show straight color banding that spinel and garnet lack.	Which species it is, whether the color came from heat or diffusion, and whether the crystal grew in the ground or in a factory. Synthetic ruby has been sold since the 1900s and looks the same in the hand.
Pink sapphire, morganite, kunzite, pink tourmaline or glass	Look for round bubbles and swirl lines with a loupe: they retire glass immediately. Kunzite fades in strong light and is often noticeably lighter than the others.	Separating the natural pinks needs refractive index and spectroscopy. Irradiation, heat and lab growth are all invisible in a photo and in the hand.
Carnelian or banded agate	Hold it against a bright lamp: chalcedony glows orange at the edges and often shows faint curved banding that the polished face hides.	Whether the color was dyed and heated, which is routine for commercial carnelian and undetectable outside a lab.
Jasper	It scratches glass and passes no light even at a thin edge, which separates it from carnelian and from most dyed lookalikes.	Whether a trade name attached to it corresponds to any real locality or species.
Sapphire, blue spinel or blue topaz	Corundum is 9 on the Mohs scale and shows almost no wear on facet edges after years of daily wear; topaz chips along one direction and shows nicks.	Heat treatment, which most commercial sapphire has had, and lattice diffusion, which colors only the surface. Both are standard, both are lab-only findings.
Tanzanite or iolite	Turn the stone slowly under one light: both show two or three distinct body colors from different directions, which sapphire does not.	Essentially all tanzanite on the market is heat-treated, and no home observation reveals it.
Lapis lazuli or sodalite	Brassy pyrite flecks under a loupe support lapis; white veining without any metallic flecks supports sodalite.	Dyeing and waxing are common on low-grade lapis and are not visible without lab testing.
Turquoise, dyed howlite or reconstituted turquoise	Acetone on a swab, applied to a hidden spot, lifts color from dyed howlite and leaves turquoise unchanged.	Stabilization with resin, which most commercial turquoise has had, and whether the material was reconstituted from powder.
Emerald	The internal garden is the signature: natural emerald is expected to be included, and a flawless bright green stone at a low price is more likely glass or synthetic.	Fracture filling with oil or resin, which the great majority of emeralds carry, and natural against synthetic origin.
Peridot	The doubling is the check: look through the table with a loupe and the back facet edges appear as clean pairs. Peridot is one of the few common gems with doubling that obvious.	Little, in fact — peridot is rarely treated. Value questions still come down to size, cut and clarity grading.
Green garnet, green tourmaline, chrome diopside or glass	Bubbles and swirls under a loupe retire glass. Tourmaline shows doubling where garnet, being singly refractive, never does.	Species identification by refractive index, and whether any of the color was added by irradiation.
Jadeite, nephrite or serpentine	Weigh a loose piece in air and then the water it displaces: jadeite gives a specific gravity near 3.34 and nephrite near 2.95, while serpentine and dyed quartz sit between 2.5 and 2.7. Round bubbles under a flashlight mean glass.	The A, B and C jade grading question — bleaching, polymer impregnation and dyeing — is entirely a laboratory matter.
Amber or copal	Amber floats in saturated salt water where glass and plastic imitations sink, and rubbed briskly on cloth it picks up small paper scraps by static.	Amber against young copal, and whether inclusions were introduced deliberately. Both are laboratory questions.
Citrine, yellow sapphire, yellow topaz or glass	Bubbles under a loupe rule out the natural stones. Density in the hand separates them further: sapphire and topaz feel heavy for their size, citrine does not.	Most yellow quartz on sale is heat-treated amethyst, and heat leaves no trace a photograph or a loupe can find.
Diamond, moissanite, cubic zirconia or white sapphire	Doubling is the fastest split: moissanite shows clearly doubled back facet edges and diamond shows none. Cubic zirconia feels distinctly heavy for its size.	Natural against laboratory-grown diamond. GIA states the two are essentially the same material, so this separation needs laboratory instruments, not a loupe.
Zircon or moissanite	Both show strong doubling. Zircon is dense and often shows worn facet edges; moissanite stays sharp and shows more rainbow fire than any diamond.	Whether a colorless zircon was heat-treated, which is routine and undetectable at home.
Glass or paste	Round bubbles, swirl lines, rounded facet junctions and scratches on the table taken together are conclusive. One of them alone is suggestive.	Nothing worth a fee. A stone that is glass needs no report.
Rock crystal quartz or colorless topaz	Both scratch glass and neither shows bubbles. Topaz is denser in the hand and cleaves on one plane, which shows as a flat internal flash.	Species by refractive index, which is where the value difference actually lives.
Obsidian	Curved glassy breaks with no crystals and no round bubbles, and thin edges pass a little light against a lamp.	Manufactured glass sold as obsidian is a common trade problem and shows bubbles under a loupe.
Black onyx or black spinel	Spinel is harder and holds facet edges; onyx is chalcedony and shows a faint edge glow against a strong lamp.	Nearly all black onyx on the market is dyed gray chalcedony, and the dye is not detectable at home.
Hematite or hematine	A streak on unglazed tile comes out red-brown for natural hematite. Manufactured hematine is strongly magnetic where natural hematite is barely attracted.	Nothing of value: hematine is an inexpensive manufactured substitute, sold openly as such.
Jet	It is fossil wood: very light, warm to the touch, and it leaves a brown streak on unglazed tile where black glass leaves none.	Jet against pressed jet or plastic imitation, which needs instruments.
Chrysoberyl cat’s eye, tiger’s eye or fiber-optic glass	Move a single light source across the dome. A natural eye opens and closes as one sharp line; fiber-optic glass shows a stiff line that never sharpens.	Chrysoberyl against quartz cat’s eye, and any dyeing of tiger’s eye.
Star sapphire, star ruby or an assembled imitation	A natural star has slightly uneven rays that move with the light source. A star that is perfectly even and stays put under a moving light is usually printed or assembled.	Natural against synthetic star corundum, which is manufactured specifically for this effect and common.
Opal, doublet, triplet or imitation	Look at the stone from the side: a solid opal shows color through its full depth, while doublets and triplets show a straight glue line and a dark backing.	Whether the opal is treated or stabilized, and its origin. Opal is also the one stone here where storage advice matters more than grading.
Labradorite or moonstone	Labradorescence appears as metallic blue and gold flashing from a specific angle; the moonstone sheen is milky and floats just under the surface as you tilt it.	Little: these are rarely treated. Glass imitations of both exist and show bubbles.
Alexandrite, color-change sapphire or color-change synthetic	Compare daylight against an incandescent bulb, not an LED. A genuine change is dramatic and repeatable; a stone that merely looks darker is not changing color.	Natural against synthetic, which is the entire value question here. Synthetic color-change corundum is sold widely and is optically convincing.

What no photo can determine

Three questions decide most of a gem’s price, and a photograph answers none of them. Origin: lab-grown ruby, sapphire and diamond are the same material as mined stones, and GIA states plainly that laboratory-grown diamonds share the chemical composition and optical properties of natural ones — separation takes lab instruments. Treatment: most commercial sapphire is heat-treated, and heat leaves traces only a microscope finds. Value: price follows carat, clarity and certificate, none of which a phone camera reads.

A photo identifies species. It does not identify provenance, treatment or price — and any app that claims otherwise is selling confidence, not identification.

If the stone came out of the ground rather than out of a ring, the value question has a different, cheaper path: the checks in how to identify valuable rocks eliminate most candidates before any lab fee, and banded finds get their own treatment in the agate guide.

Two stones get a full non-destructive sequence of their own, because imitation and treatment decide their price more than species does: how to tell if jade is real weighs a loose piece against the water it displaces, and how to tell if opal is real starts with the side view that exposes doublets and triplets.

What a photo reads from a cut stone

Color and its zoningHue, saturation and whether color sits evenly or in bands. Zoning patterns separate amethyst from glass and sapphire from synthetic corundum more often than people expect.

Luster and fireHow light returns from the surface and interior. The soapy luster of moonstone, the metallic flash of labradorite and the fire of diamond-cut stones all survive a phone photo.

Cut and wearFacet sharpness, chips and scratch patterns. A soft stone worn in a ring accumulates scratches that a hard one never shows — wear history is evidence about hardness.

Rough, uncut finds are a different problem with better home tests — start at the rock identifier for ordinary stones or the crystal identifier for points and clusters, where scratch and streak checks are safe to run on material you own.

How photo gem identification works

Loose stones, ring stones, inherited pieces: the app reads what the camera can verify and matches it against learned examples of cut and rough gems. The output is a ranked shortlist — sapphire before tanzanite before iolite — with the reasons visible, because a blue stone photo is compatible with all three and pretending otherwise is how buyers get burned.

Step 1Photograph twiceOnce on a plain white surface in daylight, once against a light source. Transparency and color zoning carry most of the signal.
Step 2Read the ranked speciesCandidates with the visual evidence for each: hue range, luster, typical cuts, and the lookalikes that share them.
Step 3Run the safe checksBubbles inside mean glass. A loupe look at facet edges — rounded edges suggest glass, crisp edges crystal. No scratch tests on set stones.
Step 4Decide if it earns a labThe app flags matches where value hangs on origin or treatment. That flag is the honest end of what photos do.

Photo first, lab when it matters

The working rule is triage. Use the key and the photo identification to sort curiosity from consequence: stones you are keeping need nothing more, and stones that might be bought, sold or insured go to a gemological lab for a report. The app comparison shows how the other identifiers handle — or skip — that same honesty.

Questions people ask

Can an app identify a gemstone from a picture?

It can name the likely species — sapphire versus tanzanite versus iolite — from color, luster and cut, and rank the candidates honestly. It cannot see whether a stone is natural or lab-grown, treated or untreated, because those are chemically near-identical and photo-invisible. Species from a photo, origin from a lab.

How can I tell if a gemstone is real at home?

Test the claim, not the stone in general. Glass impostors show round internal bubbles and rounded facet edges under a loupe; moissanite shows doubled back facet edges where diamond shows none; amber floats in saturated salt water. No home test proves natural origin, but most fakes fail one of these first checks.

Is there a free gem identifier?

The key on this page is one, and it costs nothing: color, transparency and a loupe narrow a stone to a short candidate list. Free browser uploaders do a similar first pass from a photo. For a stone you might sell or insure, a graded report from a recognized lab is the document buyers accept.

How much does professional gem identification cost?

A gemological lab identification report is typically well under the cost of being wrong about a good stone: entry-level identification services from major labs start around the price of a dinner, rising with stone size and grading detail. Check the current fee schedule of the lab you would use — prices change and vary by service.

When is a photo identification not enough?

Whenever money changes hands. Buying, selling, insuring or dividing an inheritance all need a lab report, because value hangs on origin and treatment that no photo shows. A photo identification is the right tool for curiosity, sorting a collection, and deciding which stones justify the cost of proper testing.

Sources

The queries gem identifier and gemstone identifier draw 350 US searches a month each, both at keyword difficulty 1 — the cheapest entry of the three identify intents on this site — Ahrefs Keywords Explorer, US, 2026 · checked 2026-09-01
Laboratory-grown diamonds have essentially the same chemical composition, crystal structure and optical properties as natural ones, so separating them takes laboratory instruments rather than gemological observation — GIA, Is there a difference between natural and laboratory-grown diamonds?, 2026 · checked 2026-09-02
Corundum — ruby and sapphire — measures 9 on the Mohs scale; quartz gems measure 7 and glass about 5.5 — Geology.com, Mohs Hardness Scale, 2026 · checked 2026-08-18