Imagine a world where getting high-speed internet to the middle of nowhere doesn’t involve digging trenches or launching satellites into orbit. That’s the dream Alphabet’s latest spin-off, Taara, is chasing—and it’s stepping out on its own to make it happen. Once a quirky experiment tucked inside Alphabet’s “moonshot” lab X, Taara is now an independent company, ready to take on Elon Musk’s Starlink in the race to connect the unconnected. With lasers instead of satellites, Taara’s betting on light to bridge the digital divide, and it’s already making waves.

According to Financial Times, Alphabet isn’t cutting ties completely—it’s holding onto a minority stake in Taara, while Series X Capital has jumped in with fresh funding to fuel its growth. Right now, Taara’s a lean operation with just two dozen employees, but its reach is impressive. The company’s already running projects in 12 countries, from linking up Kinshasa, the bustling capital of the Democratic Republic of Congo, to boosting bandwidth at the 2024 Coachella festival when the network got slammed by selfie-posting, TikTok-scrolling crowds. Not bad for a team that’s barely bigger than your average startup’s group chat.

Eric “Astro” Teller, the big brain behind X (and yes, that’s his real nickname—he’s the “captain of moonshots”), told the FT that spinning Taara out makes sense. “We’ve realized over time that for a good number of the things we create, there’s a lot of benefit to landing just outside of the Alphabet membrane,” he said. Translation? Taara can move faster, snag cash from the market, and cozy up to strategic investors without being tangled in Alphabet’s corporate web. It’s like letting a teenager move out of the house—they’re still family, but now they’ve got room to figure things out on their own.

Lasers, towers, and a tiny chip

So how does Taara actually work? Picture this: a traffic light-sized box perched on a tower shoots an invisible beam of light to another box miles away. That beam carries data—up to 20 gigabits per second over nearly 12.5 miles, to be exact. It’s not sci-fi; it’s optical communication, and it’s fast, cheap, and way easier than laying fiber optic cables across rivers, mountains, or islands. Need internet on a remote hillside? No problem—stick a tower up there, aim the laser, and you’re online.

The tech’s been around in some form since Taara’s days as part of Project Loon, another X experiment that tried to beam internet from giant balloons floating 20 miles above Earth. (Spoiler: balloons didn’t work out—more on that later.) But Taara’s taken the idea and grounded it—literally—by mounting its gear on towers. And just last month, the company dropped a bombshell: they’ve shrunk their tech into a compact chip. Think of it like going from a clunky desktop computer to a sleek smartphone. Taara says this chip will power a new product hitting the market in 2026, and it’s got people buzzing about what’s next.

Taking on Starlink

If Taara’s got a rival in its sights, it’s Starlink, Musk’s satellite internet juggernaut that’s already beaming broadband to rural homes and RV campers worldwide. Starlink’s approach is all about low-orbit satellites—thousands of them—blanketing the planet with coverage. Taara, on the other hand, is keeping its feet on the ground (or at least on towers). Founder Mahesh Krishnaswamy isn’t shy about the comparison. “We can offer 10, if not 100 times more bandwidth to an end user than a typical Starlink antenna, and do it for a fraction of the cost,” he told WIRED. Bold words, but Taara’s got a point: lasers don’t need rocket launches, and towers are quicker to set up than a constellation of orbiting hardware.

Still, it’s not an apples-to-apples fight. Starlink’s satellites can cover vast swaths of empty land or ocean where towers can’t reach—think ships at sea or Arctic outposts. Taara’s sweet spot is more specific: rural villages, small towns, or festival grounds where you can plop down a couple of terminals and call it a day. It’s less about replacing Starlink and more about offering a cheaper, faster alternative where it makes sense.

From balloons to beams

Taara’s story starts with failure—sort of. It was born from the ashes of Project Loon, a wild X idea that aimed to bounce internet signals between high-altitude balloons. Launched in 2013, Loon sounded cool on paper: a floating network of helium-filled orbs zapping lasers to each other, delivering data to the ground below. But reality hit hard. Keeping balloons aloft and on course was a logistical nightmare, and the costs didn’t add up. By 2021, Loon was grounded for good, just three years after it “graduated” from X as its own entity.

But Loon wasn’t a total bust. Its laser tech lived on, handed off to Krishnaswamy, who turned it into Taara’s tower-based system. Meanwhile, some of Loon’s DNA also spawned Aalyria, another Alphabet spin-off that’s working on its own laser-based project, Tightbeam, aimed at syncing up satellites and airborne networks. It’s a bit of a family tree—Taara and Aalyria are like cousins, each taking Loon’s legacy in a different direction.

What’s next?

Taara’s got big plans, but it’s still early days. That chip they’re teasing for 2026 could be a game-changer, making their tech smaller, cheaper, and easier to roll out. If they pull it off, you might start seeing Taara terminals popping up in places fiber and satellites can’t touch—think remote schools, clinics, or even disaster zones where the internet’s a lifeline. And with Alphabet’s backing (however partial) and new investors on board, they’ve got the cash to experiment.

Will Taara dethrone Starlink? Probably not anytime soon—Musk’s operation is too entrenched, with over 7,000 satellites already in orbit as of early 2025. But Taara doesn’t need to win the whole game to matter. If it can carve out a niche—delivering screaming-fast internet to rural corners at a bargain price—it might just prove there’s room for more than one player in the sky (or on the ground). For now, it’s a scrappy underdog with a laser focus—literally—and a shot at lighting up the world, one beam at a time.