SpaceX · Starlink · 7 Nov 2026

Musk says Starlink V4 lasers could outcompete undersea cables

Speaking at the Baron Investment Conference on Friday, Elon Musk said Starlink version 4 will have about ten times the capability of version 3, with next-generation laser links at 10 terabits per second. Later the same afternoon he wrote that roughly 100 lasers per satellite, each above 10 Tb/s, across more than 100,000 satellites, should be able to outcompete undersea cables. Those numbers are Musk's ambitions for the constellation backbone. They are not a finished satellite bill of materials, and SpaceX has not named a first V4 launch date.

From Alex RSpaceX · News & analysis

Updated 9 Nov 2026

The editor holds Tesla and SpaceX shares. Supercharged Daily is independent and not affiliated with either company. Nothing here is financial advice.

What Musk said at Baron, then on X

In the fireside clip shared by Sawyer Merritt, Musk described Starlink V4 as having about ten times the capability of V3, and V3 as about ten times V2, "especially with the long-distance laser links." He put next-gen laser links at 10 terabits per second ("100x better") and said SpaceX would put 10 of them on each satellite. The clip is @SawyerMerritt's post from 6 Nov 2026 at 2:10pm ET (11:10am PT).

Twenty-two minutes later, Musk posted his own numbers. In a 2:32 p.m. ET (11:32 a.m. PT) post quoting that clip, he wrote: "~100 lasers per satellite, each operating above 10 Tb/s, with over 100k satellites should be able to outcompete undersea cables." He added that the speed of light in space is about 40% faster than in cable, and that the path is direct instead of following the outline of land masses.

Ten lasers, or about a hundred?

The fireside line and the written post do not match on laser count. At Baron, Musk said ten lasers. On X he wrote "~100." Overnight he answered a reader who had multiplied ten lasers by 10 Tb/s and got 100 Tb/s per satellite. In an 6 Nov 2026 at 8:03pm ET (5:03pm PT; 7 Nov GMT) reply, Musk wrote: "If we shrink laser size, assuming 10X closer proximity, and have small gimbal angle for in-plane links, maybe we can get to 100 lasers/sat."

Read that carefully. Ten lasers is what he stated in the interview. About a hundred is a stretch goal that depends on smaller lasers, closer spacing, and limited gimbal angles for links in the same orbital plane. Until SpaceX publishes a V4 design or FCC filing that locks the count, treat "~100" as an engineering ambition Musk is working toward, not a settled satellite specification.

What 10 Tb/s actually measures

These laser figures describe links between satellites, the optical mesh that moves traffic across the constellation before it drops to a ground gateway or a user dish. They are not a promise that a home Starlink plan will download at 10 Tb/s. The separate consumer announcement names Residential 200 Mbps and Residential Max; nothing in Friday's remarks changed those retail SKUs.

Musk's undersea-cable claim rests on two physical points he stated on X: light travels about 40% faster in vacuum than in fiber, and orbital paths can cut straighter than cables that hug coastlines. Whether a mesh of more than 100,000 satellites "outcompetes" cables also depends on cost per bit, reliability, latency to specific city pairs, and how much traffic SpaceX can actually haul. Friday's posts give the ambition and the physics argument. They do not publish a side-by-side capacity table against the existing undersea network.

Where V3 already is

SpaceX has already started the generation before V4. According to TeslaNorth's report, Starship Flight 14 in September 2026 deployed 26 operational Starlink V3 satellites, each designed for about 1 Tbps of downlink capacity; the outlet says SpaceX's "Go For Orbit" documentary follows that mission. Musk did not give a first V4 launch date in the Baron remarks or in the Friday posts.

For users on the ground, the near-term upgrade path is still V3 launches and the ground kits SpaceX is shipping now. The same Friday afternoon, @Starlink said Router 4 is shipping with "Starlink V4 Kits" in the United States and Canada on Residential 200 Mbps and Residential Max. That "V4 Kit" label is consumer hardware. Router 4 has a Wi-Fi 7 chipset, according to Starlink's September announcement. It is a different product from the V4 satellite Musk described at Baron. Mixing the two will produce wrong expectations about what is flying versus what is sitting on a desk.

What is confirmed, and what is still open

Confirmed from Friday: Musk's Baron claim that V4 will have about 10× V3 capability with 10 Tb/s next-gen laser links and ten lasers in the interview framing; his written claim that ~100 lasers above 10 Tb/s across 100,000+ satellites should outcompete undersea cables; his later reply that reaching 100 lasers depends on shrinking the laser hardware; plus the separate V3 deployment context reported by TeslaNorth.

Still open: a locked V4 laser count, a first V4 launch date, FCC or ITU filings that match the ~100-laser ambition, any change to residential download plans tied to those backbone numbers, and independent measurement of how the optical mesh compares with undersea cables on cost and latency for real city pairs.

This site already covered Baron's separate comments on orbital solar data centers. That piece is about compute in orbit. This one is about how Musk wants Starlink's laser mesh to carry long-distance internet traffic.

— Alex R

Sources: Sawyer Merritt on X (6 Nov 2026); Elon Musk on X (6 Nov 2026); Elon Musk reply on X (7 Nov 2026); Starlink on X (6 Nov 2026); TeslaNorth Starlink V4 write-up (6 Nov 2026). Hero photo: Mike Lewinski / Wikimedia Commons, CC BY 2.0.