Point a thermal scope at a field and four parts decide what you see. The lens gathers heat. The sensor reads it. Processing cleans it up. And a small screen behind the eyepiece — the display — shows the result to your eye. The picture is never better than the worst part in that chain.
We score the two parts where more is simply better: the sensor and the display. Glass quality is scored too, through the lens's f-number. Lens size is different — a 25 mm and a 50 mm trade detail at distance for field of view, so the Buying Guide handles that choice. Processing goes unscored until a maker publishes something we can measure.
Lens
scoredgathers heat and sets your field of view
scored on: sharpness (f-number)
Sensor
scoredturns that heat into a picture
scored on: resolution + sensitivity (NETD)
Processing
not scored yetsharpens and steadies the raw signal
no published spec to score yet
Display
scoredthe screen you look through
scored on: density, size, and detail
Key insight: The industry sells you the sensor's resolution and stops talking. No spec sheet tells you how the display renders what the sensor captured — and 80 of the 160 scopes in this catalog fall short exactly there. That's why the score walks every link of the chain.
Every scope carries one blended score, built from the three scores this page derives:
It exists because the question every buyer actually asks — what's the best scope for my money — needs the whole field ranked on one axis. The weights are our stance, published so you can argue with it: the picture leads because the picture is what you buy a thermal for, what it's like to carry and run comes second, and the maker's written promise breaks ties.
One number can rank the field, but it can't tell you which scope is yours — near-identical scores can hide opposite scopes. These two land 4.7 apart:
Pulsar Thermion 2 LRF XL60
The bigger sensor sees farther and sharper — at 46 oz on the rail and $7,000.
$6,999.97
Nocpix RICO 3 H50R
13 oz lighter, 10 hours on a battery, and $4,499 — the picture is the tradeoff.
$4,499
Key insight: The number ranks the field; the three scores under it tell you which scope is yours. That's why The HuntGrade Score never prints alone — everywhere it appears, Image, Handling, and Warranty appear with it. And how far you shoot never enters the blend: distance gets its own read — the range-fit ratings below — for reasons the "What we don't score" section walks through.
Image is what you actually see through the scope, and the combining rule fits in one sentence: a weighted average — Resolution 40%, Sensitivity 20%, Lens 10%, Display 30%. That is the entire formula. Every scope page shows the four parts under the number, so nothing hides inside the average. Three parts score what the scope captures; the Display part — new in v4 — scores how faithfully the screen hands it to your eye.
Raw sensor pixel count (256 / 384 / 640 / 1280) — the plain number buyers already know.
Native NETD — how faint a heat difference the sensor can see. A cleaner picture when the whole field runs the same temperature.
f-number — how much heat the glass delivers to the sensor. The optical ceiling.
Screen pixels and window size — how faithfully the screen hands the picture to your eye.
40% · Resolution
The pixels on the sensor — the biggest single driver of what you see. Points run on a straight line: every doubling of pixels buys the same step, about 14 points, from 256-class at 30 to 1280-class at 98 (a 640×512 lands at 69). At today's pixel sizes every extra pixel is real captured detail, so the credit never flattens — where your eye can't use the extra pixels, the Display part of the score already catches it. We score the sensor you bought, not the zoom printed next to it: digital zoom crops the middle of the sensor and stretches it, so it never adds detail and never earns points.
20% · Sensitivity
How small a temperature difference the sensor can tell apart — its native NETD, in millikelvin, lower is better. On a damp gray night when the whole field sits within a degree of itself, this is the spec that decides whether a coyote is a shape or a smudge. We score the sensor's native figure, not the marketed one: makers often publish a processed "system" number smaller than the sensor's real spec, so we trace every claim back to the sensor inside and score that — and where the trail runs thin, we score the conservative figure.
10% · Lens
The f-number says how much of the scene's heat the glass delivers to the sensor — lower is faster, faster is better. Lens size (the millimeters) is deliberately not scored: a 25 mm and a 75 mm are different hunts, not different quality, and the Buying Guide handles that choice.
30% · Display — the part no spec sheet publishes
The display is a small screen behind the eyepiece, and no maker tells you how it performs. So we compute it from two things their own marketing implies: the screen's resolution, and the window — how big the picture looks through the eyepiece, from the marketed magnification and field of view. Three checks, blended 50 / 30 / 20:
Detail and Stretch both count pixels per degree, but they watch different layers. Stretch asks how much picture the sensor put in the window; Detail asks how finely the screen draws it. A sharp screen can draw a thin picture smoothly — no grid, still mush — and a dense picture can land on a coarse screen, where the grid shows. Different failures, and the catalog has both.
50% · Detail
Screen pixels per degree of window. Your eye resolves about 60 px/°, so 45–60 is the sweet spot. Below about 30 the screen's own pixel grid shows — the screen-door effect, like glassing through mesh. Past 60 the extra pixels are real, but your eye can't see them, so the credit flattens.
30% · Stretch
Sensor pixels per degree of window. A big window fed by a small sensor smears every sensor pixel across more of your eye — the classic case is a 384-class sensor stretched over a 35° window, which the field reliably calls mush. At 30 sensor-px/° or better there's nothing to penalize.
20% · Window
How big the picture looks through the eyepiece. A floor, not a ladder: 30° or wider is full points and going bigger earns nothing — bigger over a small sensor is how mush gets made. Under about 12° you're looking down a straw.
80 of 160 sit below the sweet spot — and not one spec sheet discloses it.
Worked example — same sensor, same glass, two displays
Two AGM builds of the same 640 core behind the same f/1.0 glass — so the first three parts are identical, and the last 30% is the whole contest. These cards read the engine's own baked numbers — the same scores you'll find everywhere else on the site.
Sub-scores are shown to one decimal; the engine blends the unrounded values and rounds once at the end, so a by-hand total can land a tenth off.
Key insight: The display is where identical sensors diverge. The Adder buys a grid-free screen and a full-size window; the Rattler keeps the same 640 pixels denser in a smaller one. Two defensible ways to build the same core — and each point of the difference has a name, which is what the old flat "display resolution" number could never show.
The bottleneck line
An average can bury a weak link — that's the price of a formula you can check yourself, and we'd rather pay it in the open than hide the rule. So every scope card carries a bottleneck line naming the weakest part of its chain, with the number that earned it. The score tells you how good the picture is; the bottleneck tells you what to blame.
The scale
Scores are anchored and absolute — not a percentage, not a curve against this year's market. An 80 means the same thing next season, because the anchors don't move. And the scale compresses on purpose: even the weakest sensor sits in a scope whose lens and screen still do part of the job, so nothing lands much below 50. Read a 52 as a usable tool, far from great — not "half as good as a 96."
HuntGrade Note: Pixel pitch — how big each sensor pixel is — sits on every spec sheet, but it isn't part of our score. Nearly every current scope uses the same 12-micron pixel, smaller isn't simply better (bigger pixels are more sensitive in bad weather), and what pitch actually changes we already score directly as resolution and sensitivity. Scoring it again would pay the same physics twice.
HuntGrade Note: Our sensitivity number may not match the NETD a brand prints. Some spec sheets publish a "system" figure — the same sensor reads better behind faster glass, so the number flatters the package. We score the sensor's native figure on every scope, and the glass separately at its own 10%; counting the lens twice would pay the same part two times.
Handling is whether you can actually live with the scope on a rifle through a long hunt. We score the four things you can measure — how much it weighs, how long the battery lasts, how weatherproof it is, and how forgiving the eye relief is — and blend them into one score, weighted by how much each matters in the field. Weight carries the most, because it's what you feel every minute you're behind the gun.
Weight leads at 60%; battery, weatherproofing, and eye relief fill in the rest.
60% · Weight
What you feel every minute the rifle's shouldered, so it carries the most. Every ounce costs the same — about a point and a half off the score per ounce — from the lightest scope on the market near the top of the scale to the heaviest near the bottom. At 60% of handling, weight alone can make or break the number.
20% · Power
Battery life, on a straight line: 4 hours sits at the bottom of the scale, and each extra hour buys the same step until 12 hours earns full marks. Swap-in batteries earn no separate points — nearly every scope takes one now, so a spare in your pocket is the plan either way; the hours between swaps are what separates scopes.
10% · Build
Weatherproofing and toughness, straight from the sealing rating. A rating that allows submersion earns more than one that only handles rain, and a maker that states no rating at all gets judged on what it does claim. It barely moves the score, because nearly every scope is already rifle-rated — so it mostly breaks ties.
10% · Eye relief
How far back you can hold the scope and still see the full picture — too short and a hard-recoiling rifle can bite you. More is more forgiving.
Worked example
Two scopes at opposite ends — and the tougher build doesn't win. Each factor's sub-score times its weight, added up:
Key insight: Weight wins handling. The toughest build in the field can't save a heavy scope — a 39 oz tank with a 4-hour battery still handles worse than a 19 oz scope that runs all night. That's why weight is 60% of the score, and why build, nearly identical across every modern scope, only breaks ties.
HuntGrade Note: We score the handling you can measure — weight, battery life, weatherproofing, and eye relief. Menu layout and button feel matter too, but they're subjective and easy to spin, so we leave them out of the number rather than guess; if a scope's interface is a real problem, that's what owner reports are for.
A thermal scope is electronics riding a hard-recoiling rifle, so coverage matters. Warranty scores the protection you get: how many years the maker backs the electronics, plus a bonus if that coverage transfers when you sell it. It's the one score that's a promise rather than a measurement.
+ 5 points if the warranty transfers to the next owner — it protects resale value, so it counts. The years top out at five, so longer coverage doesn't keep adding points.
Worked example
A long, transferable warranty versus a short one that doesn't follow the scope:
Key insight: Warranty is nearly all-or-nothing. Most makers cover the electronics for five years and let it transfer; a handful — including some of the priciest U.S. brands — cover only three and don't. There's almost nothing in between, so a long, transferable warranty is one of the clearest signals of a maker that stands behind the gear.
HuntGrade Note: This score is the years of coverage plus whether it transfers — the two things that actually affect you. A long warranty signals confidence, but it isn't a quality score: read the actual terms before you buy, because what's covered (the sensor, the housing, accidental damage) varies by maker. Where the terms carry a real catch — coverage that runs from the date of manufacture rather than purchase, or electronics-only fine print — the scope page says so in plain text next to the score.
Putting the last three sections together on a single scope, using the live engine's scores. Each score is built on its own; the blend comes last, with the arithmetic in plain sight.
RIX DBH D12
Image. A 1280 sensor — the most pixels in the field — paired with a fast lens and a clean 40 mK core puts it at 89.2 — #4 of 160 on the image board.
Handling. At 55.5 oz with a 7-hour battery, it's a haul to run all night — so handling lands well down the field at 47.
Warranty. 5 years on the electronics, and it transfers to the next owner → 95.
Key insight: The blend says top-five scope. The three scores under it say what kind: the sharpest image on the market, carried at 56 oz. If you sit a stand all night, that's your scope; if you cover ground on foot, the Handling 47 under the big number is the line that matters — which is exactly why the number never prints alone.
How far you shoot changes which scope is right, but distance never touches the HuntGrade Score. Instead, every scope carries three range-fit ratings, one for each test distance — 50, 150, and 250 yards. Each rating comes from the same two facts, both computed from the maker's own marketed numbers: how much ground the scope shows at that distance, and how much detail it puts on a hog-size animal.
- Pick a distance on Rankings and every scope's score is multiplied by its fit for that distance before ranking. A scope that fits keeps its full score; a poor fit slides down the list. Nothing is ever hidden — a weak fit is shown, not removed.
- The width side is unforgiving on purpose: showing half the ground you need at 50 yards costs far more than half the rating. A long lens that's exactly right at 250 yards can be a keyhole at 50 — and its ratings say so.
- The detail side rises with the pixels the scope puts on a hog at that distance — the same physics as the identification distances on every scope page.
- The three ratings sit in the Range fit column on Rankings, labeled by distance, with each scope's strongest distance highlighted.
| Factor | Weight / Action | What It Means |
|---|---|---|
| 50 yd | 70% width · 30% detail | Up close the job is finding and tracking a moving animal, so how much ground you can see carries most of the rating. |
| 150 yd | 50% width · 50% detail | The everyday middle: enough width to find the animal, enough detail to be sure of it. |
| 250 yd | 20% width · 80% detail | At distance the job is being certain of what you're looking at, so detail carries the rating. |
HuntGrade Note: Why not fold distance into the one number? Because there is no right answer to how far you shoot. A wide scope for timber and a narrow one for open country are each the right tool for somebody — one blended number would have to crown one kind of hunting over the rest. The fit ratings keep the score price-blind and distance-blind, and let the distance you actually shoot do the re-ranking.
Several things the spec sheet leads with never touch a HuntGrade score. They matter when you buy — but none of them makes a scope better on its own, so scoring them would mislead. We handle each where it belongs instead.
- Lens size is a tradeoff, not a winner. A bigger objective sees farther but narrows your view and adds weight and cost; a smaller one is wider and lighter. There's no better, only the right one for your country — so the Buying Guide walks you through the tradeoff instead of scoring it.
- Shot distance stays out of the blend for the same reason. How far a scope identifies is its sensor's pixels spread across its lens's field of view — and the pixels are already 40% of Image, while the lens is the tradeoff above. Score distance and you'd pay the pixels twice and crown every long lens over scopes that suit most hunts better. Distance gets its own read instead: the three range-fit ratings above, which re-order Rankings when you pick a typical shooting distance — without ever touching the score.
- Add-on features — built-in rangefinders, second lenses — earn no points for being there. Their real costs already land where they belong: the ounces in Handling, the dollars against price. Rankings marks which scopes carry a built-in rangefinder so you can filter for one.
- Price never touches a score. A scope earns its image, handling, and warranty no matter what it costs — it gets nothing for being expensive. How price lines up against capability is its own story, on The Market page.
- Digital zoom ceilings. "Up to 10×" is a software crop, not a capability — it never adds detail and never earns points anywhere on this site.
- Refresh rate. 155 of 160 scopes run 50–60 Hz, so it no longer separates anything; the rare slow ones carry a note on the scope page instead of a score.
- Pixel pitch and "system" NETD — see the notes in the Image section. Both would pay the same physics twice.
HuntGrade Note: Keeping price out of the scores is exactly what lets The Market's value read work: because the scores are price-blind, setting them against price shows which scopes are genuinely worth it — and which charge for the badge.
Two numbers do two different jobs. A score is absolute and anchored — a scope that earns an 84 today earns an 84 in ten years, and adding new scopes to the database never changes it. A rank is relative to the current field — where a scope stands against everything on the market right now — so as better scopes arrive, a rank can slip even though the score holds. The HuntGrade Score and Image both carry ranks, because those are the two reads hunters compare head-to-head.
- Score answers how good, exactly — built from the specs and fixed for good.
- Rank answers how it stacks up right now — recomputed as the field changes.
- On Rankings the rank is contextual: it reads against the whole market and narrows to the field you've filtered — #3 of 160 becomes #1 of 12 when you narrow to the long-range scopes.
HuntGrade Note: This is why a scope can hold its score for years but slowly lose its rank — the gear around it got better. The score is the honest measure; the rank is the moving target.
Every score is built from published specifications — sensor resolution, NETD, lens f-number, weight, runtime, IP rating, warranty terms — taken from the maker and cross-checked against retailer listings and owner reports. Where a maker's number looks inflated, we use the conservative figure and say so. Where a spec is thin or unverified, the score carries less confidence.
- Manufacturer specs are the baseline, verified against independent listings where possible.
- We don't extrapolate — each scope is scored on its own specs, never a sibling model's.
- Sensitivity is graded on the sensor's native NETD, not the smoothed system number some brands print.
- The window size behind the Display score is computed from the maker's own marketed magnification and field of view — the same numbers they sell on.
HuntGrade Note: The honest gap: almost nobody runs instrumented endurance testing on thermal scopes, so reliability isn't a published score yet — we're building the failure database first. And we score only the measurable parts of handling; menu feel is left to owner reports. We'd rather name a gap than fake a number.
The methodology is versioned, and any change to how a score is computed is logged here in plain sight. When a formula changes, every affected scope is recomputed — so the scores you see always match the current method.
- v5.2 (Aug 20, 2026 — live) — Two changes from one audit. First, every Handling and Warranty part now scores on one shared scale: 30 at the worst on the market, 95 at the best, a straight line between. Before this, each part ran its own range — battery life could reach 100 while the lightest scope on the market topped out around 90 — so the published weights weren't the true influence. Second, the blend moved to Image 75%, Handling 20%, Warranty 5% (was 60/30/10). Together the two fixes end a real failure: light, long-running scopes with weak sensors were out-ranking scopes with far better pictures — one $795 256-class scope sat above seventeen 640-class scopes. Now no 256-class scope out-ranks any 640-class scope. The swap-in battery bonus is retired in the same pass — nearly every scope has one now, so it stopped telling scopes apart. All 160 scopes recomputed and live.
- Range fit (Aug 8, 2026 — live) — Every scope gains three range-fit ratings — 50, 150, and 250 yards — computed from how much ground it shows and the detail it puts on a hog-size animal at each distance. Picking a distance on Rankings multiplies each score by its fit before ranking; the HuntGrade Score itself is untouched. All 156 scopes rated.
- v5.1 (Aug 7, 2026 — live) — Resolution rescored on a straight line: every doubling of sensor pixels now buys the same step, about 14 points. The old curve flattened at the top, which under-credited 1280-class sensors — at today's pixel sizes every extra pixel is real captured detail, and the Display part of the score already handles the cases where your eye can't use the extra pixels. A typical 640-class Image score drops about 5 points; class order is unchanged. All 156 scopes recomputed and live.
- v5.0 (Aug 6, 2026) — The HuntGrade Score: Image 60%, Handling 30%, Warranty 10%, blended into the one number that now leads Rankings. Shot distance stays out of the blend (it's a filter — scoring it would pay the sensor's pixels twice), add-on features earn no points, and the three scores always print beside the number. All 156 scopes scored.
- v4.0 (Jul 13, 2026) — Image rebuilt as the chain score: Resolution 40%, Sensitivity 20%, Lens 10%, and the display finally scored at 30% — Detail (screen px/°), Stretch (sensor px/°), and Window (a size floor), computed from the screen and the window the maker's own marketed numbers imply. Still a straight weighted average you can check yourself. Adds the bottleneck line on every card. Power gains +10 for a field-swappable battery (121 of 155 qualify). All 155 scopes recomputed and live.
- v3.0 (Jun 2026) — Image rebuilt as a true weighted blend — resolution 50%, sensitivity 25%, lens 15%, display 10% — scored on total pixels with diminishing returns. A standout smaller sensor can now out-score a weak bigger one. Handling weight and warranty years were both made linear.
- v2.1 (Jun 2026) — Handling reweighted so weight leads (60%); the runtime curve flattens past a full night.
- v2.0 (Jun 2026) — Image moved to a core-based model built around the sensor.
- v1.x (Jun 2026) — Letter ratings retired site-wide; every read is now a 0–100 score, and Image also carries a rank.
HuntGrade Note: Spotted a spec we got wrong? That's the fastest way to improve the data — tell us, and we'll recompute and date the change.
See the methodology in action.
Browse all rated scopes and filter by dimension to find the right scope for your priorities.
Browse all scores