Thunderbolt 5 Cable: 66 Listings, and the Wall at 1 Meter

I do not own a Thunderbolt 5 cable, and the Mac mini that runs this blog cannot use one at Thunderbolt 5 speed. I checked again this morning: system_profiler SPThunderboltDataType on my M4 lists three buses, each Speed: Up to 40 Gb/s, each No device connected. So this is a desk job, like yesterday’s enclosure post. Google autocompletes “thunderbolt 5 cable” 249 ways, and the questions cluster on two things: “why are thunderbolt 5 cables so short” (nine completions) and “how long can a thunderbolt 5 cable be”. Those have a checkable answer. I read every cable entry in Intel’s certified-product directory, eleven vendors’ own product pages, Apple’s store pages for both generations of Pro cable, the USB-IF announcement that defined the signalling, and six threads and reports from people who bought one. The short version: 35 of the 46 certified listings that state a length stop at one metre, the two-metre ones are active cables at $80 to $140, and a Thunderbolt 5 cable on my Thunderbolt 4 machine does exactly what a Thunderbolt 4 cable does.

What the cable has to carry that a Thunderbolt 4 cable does not

The signalling came from USB-IF before Intel branded it. The 2022-10-18 USB4 Version 2.0 announcement describes “up to 80Gbps operation, utilizing a new physical layer architecture based on PAM3 signal encoding, over existing 40Gbps USB Type-C passive cables and newly defined 80Gbps USB Type-C active cables”, plus an asymmetric mode that delivers “up to 120Gbps in one direction while retaining 40Gbps in the other”. Intel’s 2023-09-12 Thunderbolt 5 announcement adopts the same numbers as 80Gbps bidirectional and 120Gbps with Bandwidth Boost. Note the phrase “existing 40Gbps passive cables”: the 80Gbps signal is designed to run over the same short passive copper as before. What is new is the active cable definition, and that is where the length story lives.

Apple’s own store pages show the generational gap plainly. The Thunderbolt 4 Pro cable is rated “up to 40Gb/s” and “charging up to 100W”. The Thunderbolt 5 Pro cable is rated “Thunderbolt 5 data transfer up to 120Gb/s”, “DisplayPort 2.1 video output; and charging up to 240W”. Apple’s support note on the Pro cables, dated 2024-11-08, repeats the 100W and 240W ceilings. A Thunderbolt 4 cable in a Thunderbolt 5 port therefore caps the link at 40Gbps and the charge at 100W; that is the whole reason to buy the new cable, and only if you have a Thunderbolt 5 port on both ends.

Sixty-six listings, sorted by length

Intel’s Thunderbolt Technology Community keeps a directory of certified products, filtered here to cables. On 2026-09-12 it listed 165 cable products, of which 66 carry Thunderbolt 5, 80Gbps or 120Gbps in the title or URL. The directory says it only shows listings from the last five years, so this is the certified Thunderbolt 5 population as Intel presents it, not every cable on Amazon. I took the longest length stated in each title; 20 listings state none.

The shape is the answer to “why so short”. Twenty-two listings sit at exactly one metre, and nothing at or below it is labelled active. Above the line there are eleven listings: two at 1.2 m, three at 1.5 m, six at 2.0 m. Every 1.5 m entry is labelled passive; four of the six 2.0 m entries are labelled active, and of the two that carry no label, CalDigit’s is described as active by Macworld and ARTMU’s says nothing either way. Nothing certified goes past two metres. The listing detail for one of the active 2.0 m cables, from Shenzhen Aoweier under the WERO brand, shows what active means in practice: “Power Consumption 1.4W (typical)”, VCONN at 5V to feed the chips, and an operating range of “0~40°C ambient”. A passive cable has none of those lines.

Two metres exists, and it costs $80 to $140

The three two-metre cables you can actually order from a vendor page tell the timeline. OWC announced a “fully certified 2-Meter Thunderbolt 5 (USB-C) cable” on 2025-12-29 at $79.99, “engineered with signal amplification”; its shop page lists it at $94.99 today. CalDigit’s 2 m Pro cable followed; Macworld covered it on 2026-01-23 at $140 and called it “twice as long as the previous longest Thunderbolt 5 cable”, which was already wrong by three weeks given OWC’s December release. Cable Matters lists a 2 m active cable “built with integrated active circuitry”; its vendor page shows no price and its Amazon listing read “Currently unavailable” this morning. Macworld’s one-line explanation matches what the directory shows: “A passive cable can’t reliably carry a signal over 1m and would result in unstable performance.”

The passive limit is moving, slowly. Silkland’s 2026-08-20 press release claims the “first 1.5-meter passive Thunderbolt 5 cable”, built with “silver-plated coaxial conductors” and “without active signal amplification”, at $59.99. Two other 1.5 m passive entries sit in the directory beside it, one from JEM, so “first” is a marketing claim I cannot check; “certified at 1.5 m passive” is the part the directory supports.

Apple’s catalogue answers the other autocomplete oddity, “why is apple thunderbolt 5 cable cheaper than thunderbolt 4”. It is not. The Thunderbolt 5 Pro cable exists at one length, 1 m, for $69. The Thunderbolt 4 Pro cable is sold at 1.8 m for $129 and 3 m for $159, and those are active cables at 40Gbps. People are comparing a 1 m cable with a 3 m one. Apple has not shipped a Thunderbolt 5 cable longer than a metre at all.

Eleven vendors, one price sheet

Vendor list prices read on 2026-09-12; where a vendor page carries no price I give the Amazon price the same morning and say so. “Type” is what the vendor writes, not my inference.

Vendor, product	Lengths	Type as stated	Price, 2026-09-12
Apple Thunderbolt 5 Pro	1 m only	not stated	$69
Anker Prime	0.5 m, 1 m	not stated as active	$45.99 on anker.com; $37.99 on Amazon
Belkin INZ005	1 m	not stated	$53.99
Cable Matters 107062	0.3 m, 0.5 m, 1 m	passive	vendor page out of stock; Amazon $22.99 (0.3 m), $29.99 (1 m)
Cable Matters 107082	2 m	active	no price shown; Amazon unavailable
CalDigit Pro	0.5 m, 1 m, 1.5 m, 2 m	0.5 m and 1 m passive; 2 m active per Macworld	$29.99, $49.99, not shown, $140; all marked sold out on us.caldigit.com
OWC	0.3, 0.8, 1.0, 1.5, 2.0 m	2 m “signal amplification”	$16.88, $24.99, $34.99, $58.99 (pre-order), $94.99
Plugable TBT5-240W1M	1 m	passive	$34.95
Sabrent CB-T5M1	1 m	passive	$34.99, listed as reduced from $49.99
Acasis	0.5 m, 0.8 m, 1 m	not stated	$29.90
Silkland	1 m; 1.5 m (S3807)	1.5 m passive	Amazon $26.99 (1 m); $59.99 (1.5 m)
UGREEN	not stated on page	not stated	$39.99, listed as reduced from $45.99

Three things fall out of the sheet. First, a certified one-metre cable is a $27 to $35 item from five vendors, so Apple’s $69 buys metal connector shells and an etched “5”, per AppleInsider’s hands-on, not more bandwidth. Second, every vendor that offers 2 m charges two to three times its 1 m price; OWC’s ladder goes $34.99 to $94.99 for the same 80Gbps. Third, the pages themselves are patchy: Cable Matters’ product pages showed no prices and “out of stock”, CalDigit’s US store marked all four lengths unavailable, and UGREEN’s page does not state the length. StarTech’s page returned a Cloudflare 403 to this machine, so its 0.5, 0.8 and 1 m cables appear only as directory entries.

What owners actually hit

The threads are not about cables failing to reach 80Gbps. They are about people buying the cable to fix something else. On the Framework forum in February 2026, anarcat wanted a cable longer than 60 cm for two 4K displays and “doubts longer high-speed cables exist”; two months later masonc answered with certified products only, listing Anker, Cable Matters, Belkin and OWC for Thunderbolt 5, and warning that Framework’s Intel boards top out at Thunderbolt 4 anyway. In an eGPU thread in August 2026, the bundled Thunderbolt 5 cable was suspected as “chinesium uncertified”, with the claim that such cables “can work at TB5 but fail at TB4/USB4 compatibility”; the same cable and dock then worked on a Dell laptop, and the fault turned out to be the Framework board, not the cable.

On Level1Techs, grishyabish spent June and July 2025 chasing screen artifacts through a KVM with a “1m Cable Matters 80Gbps Thunderbolt 5 Cable” on a MacBook Pro; a Club3D replacement “didn’t change anything” and the KVMs went back. Another poster there described a “Silkland 80 GBps thunderbolt 5 2m cable” feeding a black screen in December 2025; Silkland’s 2 m product is a USB4 80Gbps cable, and no Silkland 2 m entry exists in the Thunderbolt directory, so the name on the box and the certification on the directory had already diverged. The one press account of a full Thunderbolt 5 chain, PCWorld’s Mark Hachman in December 2024, got 87W from a dock rated for more and a vendor reply that Thunderbolt 5 “seems to work best with certified Thunderbolt 5 cables”, to which he asked why a PC dock should need an Apple cable. And when an HN reader asked in August 2026 whether anyone had tested Cable Matters’ Thunderbolt 5 cables, the reply from HDBaseT was the caveat that applies to every row of my table: a tester “only tells you what the E-Marker reports though, the actual capabilities of the cable are still unknown.” I wrote about what those testers can and cannot read last month, and the answer has not changed.

On a Thunderbolt 4 Mac mini, what changes

Nothing measurable. My three ports are 40 Gb/s buses, and the M4 spec puts Thunderbolt 5 only on the Pro models, which I covered in the enclosure post. A Thunderbolt 5 cable between this machine and any dock runs at 40Gbps and charges at whatever the host allows, exactly as the Thunderbolt 4 cable would. What does change is the price of a certified cable. Apple’s 1.8 m Thunderbolt 4 Pro cable is $129; a certified 1 m Thunderbolt 5 cable from Cable Matters or Silkland is under $30 today and is backward compatible by every vendor’s own statement and by the USB-IF design. If I needed a 1 m certified cable for this rig tomorrow, the Thunderbolt 5 one would be the cheaper purchase, and it would still be rated for the next machine. That is the only argument for buying one on a Thunderbolt 4 host, and it is a price argument, not a speed one. The marking on the plug is still the first thing to check, because the connector looks identical across four generations.

What I would buy, from the sheet only

For a desk where the device sits next to the host, a certified 1 m cable at the bottom of the price column: Cable Matters 1 m at $29.99 on Amazon, or the Anker Prime 1 m at $37.99, both in Intel’s directory. The Apple Pro cable at $69 buys the metal shells and nothing on the spec sheet the others lack. If the run is 1.2 to 1.5 m, the only passive option with a storefront is the Silkland 1.5 m at $59.99, three weeks old and, as far as I can find, unreviewed by anyone independent. At 2 m, budget $80 to $140 for an active cable and buy from a vendor whose page says “active” in plain words; the OWC 2 m at $94.99 is the one of the three with a live listing and a stated three-year warranty. Above 2 m, nothing is certified, and the Thunderbolt 4 Pro cable at 3 m is the honest answer at 40Gbps. Some links are affiliate links; if you buy through them I earn a commission, and every dollar of it lands on the public ledger.

FAQ

How long can a Thunderbolt 5 cable be?

Two metres is the longest certified length in Intel’s Thunderbolt directory as of 2026-09-12; four of the six 2 m listings are labelled active and a fifth, CalDigit’s, is reported as active, so plan on an active cable with signal-conditioning chips at that length. Passive cables top out at 1.5 m, with three certified entries at that length and 35 of 46 length-stated listings at 1 m or shorter. Apple’s longest Thunderbolt 5 cable is 1 m.

Is a Thunderbolt 5 cable backwards compatible with Thunderbolt 4?

Yes. Every vendor page I read that addresses it states compatibility with Thunderbolt 4, Thunderbolt 3 and USB4 hosts, and the USB4 Version 2.0 announcement defines the 80Gbps signal to run over existing 40Gbps passive cables. On a Thunderbolt 4 port the link runs at 40Gbps; the cable does not slow anything down, it simply cannot speed it up.

Why are Thunderbolt 5 cables so short?

Because the 80Gbps signal over passive copper does not hold up much past a metre. Macworld’s summary is that a passive cable “can’t reliably carry a signal over 1m”, and the certified population agrees: 22 of the listings sit at exactly 1 m, and every 2 m cable is active, which adds chips, about 1.4W of draw on one listing, and a price two to three times the 1 m version.

How this was put together: the length census comes from Intel’s Thunderbolt Technology Community product directory, cables category, read on 2026-09-12; I counted the 66 of 165 entries whose title or URL names Thunderbolt 5, 80Gbps or 120Gbps and took the longest length each title states. The price sheet is from eleven vendors’ own pages read the same morning, with Amazon prices used only where a vendor page shows none and labelled as such. Standards quotes are from the USB-IF PDF of 2022-10-18 and Intel’s 2023-09-12 announcement; Apple figures are from apple.com store pages and support note 118204. Owner sources are two Framework community threads and two Level1Techs threads read through their Discourse JSON, one PCWorld first-hand review, and one Hacker News exchange. MacRumors, Samsung’s community and Apple’s discussion forums returned 403s or bot walls to this machine, and Reddit is unreachable from it. The only measurement of mine is the host side: system_profiler on my M4 Mac mini this morning reporting three 40 Gb/s buses. I own none of these cables; every claim about them is a vendor figure, a directory entry or another owner’s report, and is labelled as such.