For current new listings, the cheapest external desktop HDD is 2.9% cheaper per TB than the cheapest bare internal SATA HDD. Market Value compares current family peers, while Historical Value checks each exact product condition against stored observations. We have not found a maker document naming the drive inside, so we do not claim it.
Cheapest in-stock external desktop HDD against the cheapest in-stock bare internal SATA drive, at the same condition. Our own numbers, computed when this page loaded.
| Condition | Best external | Best bare internal | Gap | Cheaper |
|---|---|---|---|---|
| NEW | $28.33/TB Seagate Expansion 24TB Portable HDD (24TB) | $29.17/TB Ultrastar 12TB SATA HUH721212ALE601 (12TB) | 2.9% | External enclosure |
| USED | $26.64/TB Seagate Expansion 8TB Desktop HDD (8TB) | $20.00/TB Ultrastar 10TB SATA HUH721010ALE600 (10TB) | 33.2% | Bare internal drive |
These are different products and the comparison is only about cost per terabyte. A bare drive goes straight into a bay or an array; an external is a sealed unit with a USB controller, a power supply on the desktop sizes, and a drive inside that the maker does not identify.
18 in-stock listings · 3.5in drives with their own power adapter
Best cost per terabyte here right now: Seagate Expansion 8TB Desktop HDD at $26.64/TB ($213 for 8TB, desktop hdd, used listing). We have not found a maker document naming the drive inside it.
| Model | Capacity | Market Value | Historical Value | Interface | Inside | Warranty | Condition | $/TB | Price | Buy |
|---|---|---|---|---|---|---|---|---|---|---|
| Seagate Expansion 8TB Desktop HDD | 8TB | Excellent | GOOD PRICE | USB 3.0 | NOT SOURCED | — | USED | $26.64 | $213 | Buy on Amazon |
| WD Elements 20TB Desktop HDD | 20TB | GOOD PRICE | USB 3.0 | NOT SOURCED | — | USED | $30.67 | $613 | Buy on Amazon | |
| Seagate One Touch 8TB Desktop HDD | 8TB | Good | FAIR PRICE | USB-C | NOT SOURCED | — | USED | $34.44 | $275 | Buy on Amazon |
| WD Elements 10TB Desktop HDD | 10TB | FAIR PRICE | USB 3.0 | NOT SOURCED | — | NEW | $34.55 | $346 | Buy on Amazon | |
| WD My Book 22TB Desktop HDD | 22TB | HIGH PRICE | USB 3.0 | NOT SOURCED | — | NEW | $36.71 | $808 | Buy on Amazon | |
| Seagate Expansion 22TB Desktop HDD | 22TB | FAIR PRICE | USB 3.0 | NOT SOURCED | — | USED | $36.77 | $809 | Buy on Amazon | |
| WD Elements 14TB Desktop HDD | 14TB | Average | FAIR PRICE | USB 3.0 | NOT SOURCED | — | USED | $40.14 | $562 | Buy on Amazon |
| WD My Book 6TB Desktop HDD | 6TB | Excellent | FAIR PRICE | USB 3.0 | NOT SOURCED | — | USED | $40.73 | $244 | Buy on Amazon |
| WD My Book 8TB Desktop HDD | 8TB | Average | HIGH PRICE | USB 3.0 | NOT SOURCED | — | USED | $41.61 | $333 | Buy on Amazon |
| WD Elements 16TB Desktop HDD | 16TB | HIGH PRICE | USB 3.0 | NOT SOURCED | — | USED | $41.75 | $668 | Buy on Amazon | |
| WD Elements 8TB Desktop HDD | 8TB | Average | GOOD PRICE | USB 3.0 | NOT SOURCED | — | USED | $42.16 | $337 | Buy on Amazon |
| Seagate Expansion 4TB Desktop HDD | 4TB | Excellent | FAIR PRICE | USB 3.0 | NOT SOURCED | — | USED | $42.25 | $169 | Buy on Amazon |
| WD Elements 20TB Desktop HDD | 20TB | FAIR PRICE | USB 3.0 | NOT SOURCED | — | NEW | $43.41 | $868 | Buy on Amazon | |
| Seagate Expansion 4TB Desktop HDD | 4TB | Excellent | In progress | USB 3.0 | NOT SOURCED | — | USED | $43.99 | $176 | Buy on Amazon |
| WD My Book 16TB Desktop HDD | 16TB | FAIR PRICE | USB 3.0 | NOT SOURCED | — | NEW | $46.06 | $737 | Buy on Amazon | |
| WD My Book 4TB Desktop HDD | 4TB | Expensive | FAIR PRICE | USB 3.0 | NOT SOURCED | — | USED | $51.94 | $208 | Buy on Amazon |
| WD Elements 24TB Desktop HDD | 24TB | HIGH PRICE | USB 3.0 | NOT SOURCED | — | NEW | $52.14 | $1251 | Buy on Amazon | |
| WD My Book 4TB Desktop HDD | 4TB | In progress | USB 3.0 | NOT SOURCED | — | NEW | $81.25 | $325 | Buy on Amazon |
Market Value compares current listings in the same external-drive family and condition. Historical Value uses stored observations for the exact product condition in the background. For “Inside” we have not found a maker document naming the drive in any of these products, and the contents vary by batch, so we state that rather than guessing — see what the community has found by opening them. Warranty shows a dash because the makers' warranty pages did not resolve through our source-verification path; read the term on the maker's page for the exact model. Interface is the maker's own specification string, not a speed measurement.
20 in-stock listings · 2.5in drives, USB powered
Best cost per terabyte here right now: Seagate Expansion 24TB Portable HDD at $28.33/TB ($680 for 24TB, portable hdd, new listing). We have not found a maker document naming the drive inside it.
| Model | Capacity | Market Value | Historical Value | Interface | Inside | Warranty | Condition | $/TB | Price | Buy |
|---|---|---|---|---|---|---|---|---|---|---|
| Seagate Expansion 24TB Portable HDD | 24TB | GREAT PRICE | USB 3.0 | NOT SOURCED | — | NEW | $28.33 | $680 | Buy on Amazon | |
| WD My Passport 5TB Portable HDD | 5TB | Excellent | GOOD PRICE | USB 3.1 | NOT SOURCED | — | USED | $33.59 | $168 | Buy on Amazon |
| WD Elements 5TB Portable HDD | 5TB | Excellent | FAIR PRICE | USB 3.2 GEN 1 | NOT SOURCED | — | USED | $35.75 | $179 | Buy on Amazon |
| Easystore 4TB USB External | 4TB | Excellent | GREAT PRICE | USB | NOT SOURCED | 2 yr | USED | $36.25 | $145 | Buy on Amazon |
| Seagate Expansion 24TB Portable HDD | 24TB | In progress | USB 3.0 | NOT SOURCED | — | USED | $37.46 | $899 | Buy on Amazon | |
| Seagate Expansion 20TB Portable HDD | 20TB | FAIR PRICE | USB 3.0 | NOT SOURCED | — | NEW | $37.50 | $750 | Buy on Amazon | |
| Seagate One Touch 5TB Portable HDD | 5TB | Excellent | GOOD PRICE | USB 3.0 | NOT SOURCED | — | NEW | $38.10 | $191 | Buy on Amazon |
| WD My Passport 5TB Portable HDD | 5TB | Good | GREAT PRICE | USB 3.1 | NOT SOURCED | — | USED | $38.33 | $192 | Buy on Amazon |
| Easystore 5TB USB External | 5TB | Good | GREAT PRICE | USB | NOT SOURCED | 2 yr | USED | $38.78 | $194 | Buy on Amazon |
| Seagate Expansion 6TB Portable HDD | 6TB | Excellent | FAIR PRICE | USB 3.0 | NOT SOURCED | — | NEW | $39.17 | $235 | Buy on Amazon |
| WD My Passport 6TB Portable HDD | 6TB | Good | FAIR PRICE | USB 3.1 | NOT SOURCED | — | USED | $40.00 | $240 | Buy on Amazon |
| WD My Passport 4TB Portable HDD | 4TB | Average | GOOD PRICE | USB 3.1 | NOT SOURCED | — | USED | $42.09 | $168 | Buy on Amazon |
| WD My Passport 5TB Portable HDD | 5TB | Average | FAIR PRICE | USB 3.1 | NOT SOURCED | — | USED | $43.28 | $216 | Buy on Amazon |
| WD My Passport 4TB Portable HDD | 4TB | Average | GOOD PRICE | USB 3.1 | NOT SOURCED | — | USED | $43.36 | $173 | Buy on Amazon |
| WD My Passport 6TB Portable HDD | 6TB | Average | FAIR PRICE | USB 3.1 | NOT SOURCED | — | USED | $43.89 | $263 | Buy on Amazon |
| WD Elements 6TB Portable HDD | 6TB | Average | GOOD PRICE | USB 3.2 GEN 1 | NOT SOURCED | — | NEW | $44.83 | $269 | Buy on Amazon |
| WD My Passport 6TB Portable HDD | 6TB | Average | FAIR PRICE | USB-C | NOT SOURCED | — | USED | $44.86 | $269 | Buy on Amazon |
| Seagate One Touch 5TB Portable HDD | 5TB | Average | In progress | USB | NOT SOURCED | — | USED | $48.00 | $240 | Buy on Amazon |
| WD My Passport 5TB Portable HDD | 5TB | Average | FAIR PRICE | USB 3.1 | NOT SOURCED | — | USED | $48.13 | $241 | Buy on Amazon |
| WD_BLACK 6TB Portable HDD | 6TB | Average | GREAT PRICE | USB | NOT SOURCED | — | USED | $48.42 | $291 | Buy on Amazon |
Market Value compares current listings in the same external-drive family and condition. Historical Value uses stored observations for the exact product condition in the background. For “Inside” we have not found a maker document naming the drive in any of these products, and the contents vary by batch, so we state that rather than guessing — see what the community has found by opening them. Warranty shows a dash because the makers' warranty pages did not resolve through our source-verification path; read the term on the maker's page for the exact model. Interface is the maker's own specification string, not a speed measurement.
20 in-stock listings · interface shown as the maker states it
Best cost per terabyte here right now: SanDisk Extreme 2TB Portable SSD at $120.00/TB ($240 for 2TB, portable ssd, used listing). We have not found a maker document naming the drive inside it.
| Model | Capacity | Market Value | Historical Value | Interface | Inside | Warranty | Condition | $/TB | Price | Buy |
|---|---|---|---|---|---|---|---|---|---|---|
| SanDisk Extreme 2TB Portable SSD | 2TB | Excellent | FAIR PRICE | USB | NOT SOURCED | — | USED | $120.00 | $240 | Buy on Amazon |
| External SSD 2TB | 2TB | Excellent | In progress | USB | NOT SOURCED | 5 yr | NEW | $125.00 | $250 | Buy on Amazon |
| SanDisk Extreme 2TB Portable SSD | 2TB | Good | FAIR PRICE | USB 3.2 GEN 2 | NOT SOURCED | — | USED | $136.37 | $273 | Buy on Amazon |
| WD My Passport 1TB Portable SSD | 1TB | Good | GREAT PRICE | USB | NOT SOURCED | — | USED | $139.04 | $139 | Buy on Amazon |
| SanDisk Extreme 2TB Portable SSD | 2TB | Good | FAIR PRICE | USB | NOT SOURCED | — | NEW | $142.50 | $285 | Buy on Amazon |
| Samsung T Series 2TB Portable SSD | 2TB | Good | GREAT PRICE | USB 3.2 GEN 2 | NOT SOURCED | — | USED | $144.50 | $289 | Buy on Amazon |
| Crucial X Series 1TB Portable SSD | 1TB | Good | GOOD PRICE | USB-C | NOT SOURCED | — | USED | $152.99 | $153 | Buy on Amazon |
| Samsung T Series 4TB Portable SSD | 4TB | Good | FAIR PRICE | USB 3.2 GEN 2 | NOT SOURCED | — | USED | $157.50 | $630 | Buy on Amazon |
| Samsung T Series 2TB Portable SSD | 2TB | Good | GREAT PRICE | USB 3.2 GEN 2 | NOT SOURCED | — | USED | $162.00 | $324 | Buy on Amazon |
| Crucial X Series 1TB Portable SSD | 1TB | Average | FAIR PRICE | USB 3.2 | NOT SOURCED | — | NEW | $169.14 | $169 | Buy on Amazon |
| Samsung T Series 2TB Portable SSD | 2TB | Average | GOOD PRICE | USB 3.2 GEN 2 | NOT SOURCED | — | USED | $174.53 | $349 | Buy on Amazon |
| Samsung T Series 2TB Portable SSD | 2TB | Average | FAIR PRICE | USB 3.2 GEN 2X2 | NOT SOURCED | — | USED | $175.16 | $350 | Buy on Amazon |
| Samsung T Series 2TB Portable SSD | 2TB | Average | GOOD PRICE | USB 3.2 GEN 2X2 | NOT SOURCED | — | USED | $179.50 | $359 | Buy on Amazon |
| Samsung T Series 2TB Portable SSD | 2TB | Average | GOOD PRICE | USB 3.2 GEN 2 | NOT SOURCED | — | USED | $180.00 | $360 | Buy on Amazon |
| SanDisk Extreme 2TB Portable SSD | 2TB | Average | GREAT PRICE | USB | NOT SOURCED | — | USED | $180.03 | $360 | Buy on Amazon |
| Crucial X Series 2TB Portable SSD | 2TB | Expensive | In progress | USB 3.2 | NOT SOURCED | — | NEW | $182.65 | $365 | Buy on Amazon |
| Samsung T Series 1TB Portable SSD | 1TB | Average | FAIR PRICE | USB 3.2 GEN 2 | NOT SOURCED | — | USED | $189.99 | $190 | Buy on Amazon |
| Samsung T Series 1TB Portable SSD | 1TB | Average | GREAT PRICE | USB 3.2 GEN 2X2 | NOT SOURCED | — | USED | $195.00 | $195 | Buy on Amazon |
| Samsung T Series 4TB Portable SSD | 4TB | Average | FAIR PRICE | USB 3.2 GEN 2 | NOT SOURCED | — | USED | $195.68 | $783 | Buy on Amazon |
| SanDisk Extreme 1TB Portable SSD | 1TB | Average | GOOD PRICE | USB | NOT SOURCED | — | USED | $195.81 | $196 | Buy on Amazon |
Market Value compares current listings in the same external-drive family and condition. Historical Value uses stored observations for the exact product condition in the background. For “Inside” we have not found a maker document naming the drive in any of these products, and the contents vary by batch, so we state that rather than guessing — see what the community has found by opening them. Warranty shows a dash because the makers' warranty pages did not resolve through our source-verification path; read the term on the maker's page for the exact model. Interface is the maker's own specification string, not a speed measurement.
New listings on both sides. Only capacities where both an external and a bare internal drive are in stock appear at all.
| Capacity | Cheapest external | Cheapest bare drive | Difference | Verdict |
|---|---|---|---|---|
| 1TB | $110 Seagate Expansion 1TB Portable HDD | $99 BarraCuda 1TB ST1000DM010 | +$11 | Bare drive is cheaper — do not shuck |
| 2TB | $132 WD Elements 2TB Portable HDD | $110 BarraCuda 2TB ST2000DM001 | +$22 | Bare drive is cheaper — do not shuck |
| 4TB | $195 Seagate Expansion 4TB Desktop HDD | $195 WD Red Plus 4TB WD40EFZZ | −$0 | External is cheaper before you open it |
| 6TB | $235 Seagate Expansion 6TB Portable HDD | $249 BarraCuda 6TB ST6000DM003 | −$14 | External is cheaper before you open it |
| 8TB | $333 Seagate Expansion 8TB Desktop HDD | $280 BarraCuda 8TB ST8000DMZ04 | +$53 | Bare drive is cheaper — do not shuck |
| 10TB | $346 WD Elements 10TB Desktop HDD | $372 Ultrastar 10TB SATA HUH721010ALN600 | −$26 | External is cheaper before you open it |
| 16TB | $682 WD Elements 16TB Desktop HDD | $711 SkyHawk AI 16TB ST16000VE000 | −$29 | External is cheaper before you open it |
| 20TB | $750 Seagate Expansion 20TB Portable HDD | $640 20TB SATA 7200RPM NAS | +$110 | Bare drive is cheaper — do not shuck |
| 22TB | $808 WD My Book 22TB Desktop HDD | $952 Red Pro 22TB NAS WD221KFGX | −$144 | External is cheaper before you open it |
| 24TB | $680 Seagate Expansion 24TB Portable HDD | $1265 Exos X24 24TB SATA | −$585 | External is cheaper before you open it |
A cheaper external is where the case for opening one begins, not where it ends. You do not know which drive is inside before you open it, opening it ends the warranty, and some units use a 3.3V power-disable pin that stops the drive spinning up in a normal desktop. The full shucking guide covers all three.
External storage is where fake-capacity fraud concentrates. The mechanism is simple: a small, cheap flash chip is paired with a controller reflashed to report a capacity it does not have. The drive mounts, shows the advertised size, and accepts files until it passes the real capacity, at which point new writes silently overwrite old data or return corrupt on read. The failure is designed to surface long after the return window.
Both filters run when we seed the catalogue and again when a page renders, because a catalogue is only as clean as its last refresh. If you buy outside our tables, write a full drive of real data and read it back before you trust it — where to buy safely covers the channels and their risks, and drive health covers verifying what arrived.
External and bare-drive pricing can cross in either direction because the products move through different retail channels. A periodic external-drive promotion can make the enclosure cheaper per terabyte, while another capacity can favor the bare drive.
The live comparison above computes the answer from current qualifying listings. It reports new and non-new conditions separately and does not rely on a remembered market rule.
A non-new external and a non-new bare drive are also different purchases. The bare drive may expose an exact enterprise part number, while for the external unit we have not found a maker document naming its internal mechanism. The condition column prevents that distinction from disappearing inside one average.
If the storage will remain on a desk and connect over USB, compare current external pricing directly. If it will go into a chassis or array, compatibility, identity, recording method, workload guidance, and warranty can outweigh a small price-per-terabyte difference.
Where bare drives remain clearly correct is anything going into a bay. An array wants a drive with a published model number, a stated workload rating, a warranty you can claim without having opened a case, and a recording technology you can look up before you buy. An external gives you none of those, which is the subject of the next section.
Every table on this page has a column headed “Inside”, and every row in it reads “not sourced”. That is the single most important thing on this page, and it is a deliberate refusal rather than missing data.
We have not found a manufacturer document naming the bare drive inside any external enclosure. Nothing we have found from Western Digital says whether an Elements Desktop contains a particular Ultrastar or a particular consumer drive, and nothing we have found from Seagate says what is inside an Expansion. More importantly, the contents vary between production batches of the same model number and the same part number. Two units with identical labels, bought a year apart or from different warehouses, can hold different drives.
This matters because a meaningful share of external drives contain shingled recording. That is a reasonable engineering choice for the role most externals actually serve — occasional backups, media libraries, files written once and read many times — and it is close to disqualifying for the roles people repurpose externals into, especially arrays. If you are buying an external to use as an external, the recording technology is largely academic. If you are buying one to open, it is the whole question, and you cannot answer it before you open it.
Plenty of guides state the contents as fact. They are reporting community teardowns, usually accurately for the batch that was opened, and then presenting a batch observation as a product specification. We link that work rather than repeat it as fact: our shucking guide sets out what people have found by opening these units, framed as community reports that vary by batch, which is what they are.
The honest version of the advice is short. Buy an external for the job an external does and the contents do not matter. Buy one to open and you are buying a lottery ticket with good odds, not a specification — and our CMR versus SMR list covers the bare drives whose recording technology the maker does publish, if that is the thing you actually need.
The word “external” covers three products that share a USB port and almost nothing else, and the tables above are split accordingly because ranking them together would be meaningless.
Desktop HDDs compete for the lowest storage cost per terabyte among devices that connect directly over USB. They are 3.5-inch drives in a case with their own power adapter, they reach the largest capacities in the category, and they are not portable in any practical sense: they need a mains socket and they should not be carried around while running. For bulk storage that lives on a desk — a media library, a backup target, or an archive — use the live table rather than assuming a fixed price advantage over a bare internal drive.
Portable HDDs are 2.5-inch drives that run on USB power alone. They cost more per terabyte than the desktop units and top out at lower capacities, and what you are paying for is the absence of a power brick. For a drive that goes in a bag, that difference is the entire product. The capacity ceiling here is genuinely lower and it is a physical constraint rather than a market one.
Portable SSDs are a different category wearing the same word. They cost several times more per terabyte than either mechanical family — the tables above will show you exactly how much more today — and prices in this family roughly doubled through the NAND shortage, a movement widely reported across the storage trade press during 2025 and 2026. What the money buys is stated as design facts rather than performance claims: no moving parts to survive a drop, a fraction of the physical size and weight, and no spin-up delay. We do not publish speed comparisons, and the interface column carries the maker's own specification string rather than our interpretation of it.
The practical rule that falls out of this: pick the family from where the drive will live and how it will move, then optimise cost per terabyte inside that family. Choosing across families on price alone produces the wrong answer in both directions — a desktop unit bought for travel, or an SSD bought for an archive that never leaves a shelf.
Externals are the part of the storage market where outright fraud is common rather than occasional, and it is worth understanding the mechanism because it defeats the checks most buyers run.
A fake-capacity drive is a small, genuine flash chip paired with a controller reflashed to report a size it does not have. Plug it in and the operating system sees the advertised capacity. Copy files and they copy. Check the properties and the free space looks right. The deception only surfaces once you write past the real capacity, at which point the drive either silently wraps and overwrites what you already put there or returns unreadable data. Both failures appear weeks or months after purchase, which is precisely the design.
The two filters we apply are described in the block above and both run twice, at seeding and again when a page renders. The plausibility floor is arithmetic: no supply chain delivers a terabyte of mechanical storage under eight dollars or a terabyte of flash under twenty-five, so a listing beneath those rates is making a capacity claim rather than offering a capacity. The brand allowlist handles the rest, because above 2TB the fraud is essentially all from names with no history outside the listing itself.
In our most recent discovery sweep those two filters rejected three listings that had otherwise passed every check: two portable units and a 16TB desktop drive from a brand we do not carry at that size. We list what we rejected and why rather than quietly dropping it, because a filter you cannot inspect is just an assertion.
If you buy outside our tables — and there are good reasons to, since we price one venue — the verification is straightforward and worth the hour. Write a full drive's worth of real data, read it all back, and compare. Any tool that fills the drive and verifies will do. Our where to buy coverage sets out which channels carry which risks.
Shucking — buying an external and removing the drive for use in a machine or an array — is the reason a lot of people arrive at pages like this one. The arithmetic behind it is on this page rather than in an anecdote.
The comparison that matters is not external-versus-bare in general. It is external-versus-bare at the specific capacity you want, on new listings, today. That is what the shuck table computes: the cheapest external at each capacity beside the cheapest bare internal SATA drive at the same capacity, with the difference stated in dollars. Some capacities show a real saving. Others show the bare drive costing less, and at those capacities opening an external costs you money and voids a warranty in exchange for nothing.
Even where the saving is real, the mechanism is not known until the case is open and contents can vary by batch. Disassembly can also affect support or warranty handling, so check the exact terms before opening the enclosure. Some units carry a 3.3V power-disable pin that can prevent spin-up with an incompatible desktop power connection.
None of that makes shucking a bad idea. It makes it a trade rather than a free saving, and the size of the saving is the input that decides whether the trade is worth it. When the difference is a hundred dollars a drive across an eight-bay build, most people take the trade. When it is six dollars, almost nobody should.
Our shucking guide covers the mechanics, the 3.3V pin and what the community has found inside various models, framed as the batch-varying reports they are. Our cheapest per TB table is where to look if you want the whole market ranked by rate rather than by category.
Most external drives are bought for backup, and most external drive backups are worse than their owners think — not because the drive is bad, but because one drive is not a backup strategy.
The framing worth knowing is 3-2-1: three copies of anything you care about, on two different kinds of media, with one copy off-site. An external drive is an excellent second copy and an excellent off-site copy. It is a poor only-copy, and a drive that lives permanently plugged into the machine it backs up is not really a second copy at all, because most of the things that destroy the original — theft, fire, a power event, ransomware that walks every mounted volume — take the backup with it.
That points at a specific and cheap discipline: two externals, rotated. One is connected and receiving backups; the other is somewhere else entirely, at an office, at a relative's house, in a car. You swap them on a schedule you will actually keep, which for most people is monthly. This costs one extra drive and converts a single point of failure into a genuine off-site copy, and it is the single highest-return thing most people reading this page could do with their storage budget.
On sizing, a backup target should leave room for retained versions and growth. Compare adjacent capacities in the live table rather than assuming the next size is always cheaper per terabyte.
For the machines that generate the data rather than store the copies, our NAS coverage and the drive health section cover the other half of the problem: knowing when a drive is failing before it takes something with it.
Everything here is computed when the page loads, from our own tracking of Amazon US listings, refreshed about once a day. The timestamp at the top is the moment of the last successful refresh rather than the moment you opened the page, and the distinction matters: a page claiming to be live while stamping itself with the current time is telling you nothing about the data behind it.
Out-of-stock listings are excluded entirely rather than shown at their last known price. That keeps the rankings honest at the cost of making the tables shorter and occasionally making a capacity disappear, which is the correct trade — a price you cannot pay is not a price.
Cost per terabyte is the column to sort on and it is computed from the price you would actually pay divided by the capacity you would actually receive. Neither input is under a marketer's control, which is what makes it more useful than a discount percentage. A discount is a relationship between two numbers and the seller sets one of them; raise the reference price and the badge grows without the price moving.
Listings priced far above their family's typical rate carry a visible flag and are excluded from the winner callouts and from every aggregate on this page. They are real listings so we show them, but letting a seller asking several times the going rate sit unmarked next to normal prices, or letting one drag a median, would misrepresent the market.
Condition is badged on every row. Used external listings are legitimate purchases at the right price for the right role, and they are a different product from a new one — which is why every comparison on this page pairs like with like rather than flattering one side.
Out of stock on Amazon, or no longer listed. Each page keeps its specifications and points to what is in stock now; none of them has a current price.
136 product pages, 100 with a live offer today. Each page carries the listing's specifications, its condition and provenance, and in-stock alternatives. Prices are on the tables above, not in this index.
It changes with capacity, condition, and current inventory. The live comparison above computes the cheapest qualifying external desktop drive against the cheapest qualifying bare internal SATA drive in the same condition. It shows both condition tracks rather than asserting that one format is always cheaper.
We have not found a maker document naming it, and community teardowns show the contents vary between production batches of the same model number. That is why every table here reads “not sourced” rather than naming a drive. The community documents what it finds by opening them, and our shucking guide links that work as the batch-varying reports it is. Any guide stating the contents as a specification is asserting something the manufacturer does not.
Start with arithmetic. We reject any external HDD under $8/TB and any external SSD under $25/TB, because nothing legitimate is manufactured and shipped below those rates. Then check the brand: above 2TB, fake capacity is essentially all from names with no history outside the listing. A pocket-sized “16TB portable” needing no power adapter does not exist at consumer prices. If you buy outside our tables, write a full drive of real data and read it back before trusting it.
It depends on the live gap at the required capacity, which the shuck table computes. Even where the saving is real, the internal mechanism cannot be confirmed before opening, support or warranty handling may change after disassembly, and some units use a 3.3V power-disable pin that requires a compatible power path.
Decide by where the drive lives, then compare the live rates. Desktop units are 3.5-inch drives with their own power adapter and usually offer the largest capacities in this category. Portable units run on USB power alone and trade capacity density for easier transport. Pick the family first, then optimise cost per terabyte inside it.
Flash and mechanical storage are different manufacturing economics, and the gap widened sharply through the NAND shortage — portable SSD prices roughly doubled over that period, as widely reported across the storage trade press. The tables above show today's exact spread between the families. What the premium buys is stated as design facts rather than speed claims: no moving parts, a fraction of the size and weight, and no spin-up delay.
Because the makers' warranty pages did not resolve through our source-verification path, and an Amazon listing's warranty claim is not a manufacturer specification. Rather than print a figure we could not verify, the column shows a dash. Read the term on the maker's own page for the exact model you are buying — the periods do genuinely differ between the value lines and the premium ones.
Not as an external, and opening one to get the drive out is the shucking question above. Arrays want a drive with a published model number, a stated workload rating, a claimable warranty and a recording technology you can look up before buying — an external gives you none of those, and a shingled drive in an array can extend a rebuild dramatically. Our NAS hard drives coverage lists the drives that publish all four.
Specification claims on this page were checked against the pages below on 4 August 2026. Where a widely-repeated figure did not survive that check, we report what the source says now and say so in the text rather than repeating the familiar number.
Wondering which way prices are going? The external drive price forecast reads desktop HDD, portable HDD and portable SSD separately, because averaging them measures the product mix rather than the market.