Mecka AI nears $500M valuation in Sequoia-led deal amid rush for robot training data
TechCrunch Marina Temkin
Mecka AI is closing in on a $500M valuation in a Sequoia-led round. The bet is that robot-makers are now fighting over training data, just like AI labs did.
Based on reporting by TechCrunch, Marina Temkin — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
Mecka AI is heading toward a new Sequoia Capital-led round at a valuation of about $500 million, according to two people familiar with the deal. The startup only closed a $60 million round three months ago, so this is a fast jump even by AI-era standards. The new financing isn’t final yet, and the size of the round still hasn’t been nailed down.
The company sells something a lot of robot builders now seem to want badly: real human motion data. Mecka pays people to record ordinary tasks such as making coffee or fixing cars, using body sensors and smartphones to capture how bodies move in the real world. Its pitch is simple enough. If robots are going to work outside a lab, they need data that looks like the real world, not just clean simulations and tidy demos.
That idea is the whole business. Mecka was founded in 2024 by four entrepreneurs who don’t come from robotics, but who saw a shortage of physical-world data and a bottleneck around capturing real interactions. Two of the founders, Josh Gao and Mogen Cheng, are Canadians who previously built a restaurant fintech startup. Jason Chong later joined Coinbase after Coinbase bought his crypto exchange, and Duy Nguyen handles operations. The company takes its name from “mecha,” the giant robot genre.
The comparison Mecka wants is obvious: Scale AI, Mercor, Surge, and other companies that built businesses around human data for large language models. Now the same playbook is moving into robotics. Mecka hasn’t disclosed customers, but the broader market is already leaning on real-world capture methods like this “egocentric” approach, along with teleoperation and other physical data collection. And the money is arriving fast. Last week, TechCrunch reported that XDOF was also nearing a new round, at a $1.2 billion valuation.
In early June, Gao told Fortune that Mecka was projecting a $100 million annual run rate by the end of 2026. That is the kind of number investors hear and start reaching for the chequebook. But it also says something slightly uncomfortable about the sector: robots still need humans to teach them how to move, and the people selling the training data are starting to look like the real winners.
My take — AI-written commentary, not fact-checked reporting
This is the same old AI pattern wearing a robot costume: data suppliers rush to the front of the line and call it infrastructure. Sequoia is basically betting that the bottleneck in humanoids is not metal arms or flashy demos, but the boring human footage nobody wanted to collect until now. That’s usually where the money is, right before everyone pretends it was obvious.
Read more about this at: TechCrunch
