Mecka AI, a two-year-old startup that pays people to record everyday movements so machines can learn from them, is nearing a new funding round led by Sequoia Capital at a valuation of about $500 million, according to two people with knowledge of the deal. The talks come roughly three months after the company announced a $60 million raise led by Framework Ventures, with checks from Menlo Ventures, SV Angel, and Kindred Ventures.
The precise size of the new round has not been confirmed, and the terms are not final, so the figure could still change. Mecka AI did not respond to a request for comment, and Sequoia declined to comment.
Also read: Pocket FM Doubles Annualized Revenue to $500M as AI Now Drives 93% of Its Catalog
From crypto exchanges to motion capture
Mecka was co-founded in 2024 by four entrepreneurs who, by their own account, have no robotics backgrounds. Canadians Josh Gao and Mogen Cheng previously built a restaurant fintech startup; Jason Chong joined Coinbase after it acquired his crypto exchange. The fourth co-founder, Duy Nguyen, the only non-Canadian on the team, runs operations.
What the group spotted was a gap rather than a product. General-purpose robots, humanoids included, were failing not because of weak model architectures but because there was not enough data describing how the physical world actually behaves. Capturing real-world interactions, they concluded, was the binding constraint.
Also read: Anthropic's Rogue Mythos 5 AI Broke Out of Its Sandbox—Then Got Stuck on CAPTCHA for 150 Pages
The company’s name nods to “mecha,” the fictional giant robot piloted by a human. Its business model borrows from the human-data vendors that rose alongside large language models: pay contributors to generate labeled data, then sell it to the labs training the next generation of models. Where Scale AI turned to annotators for text and images, Mecka turns to contributors wearing body sensors and carrying smartphones.
The economics of egocentric data
Contributors record themselves doing ordinary things, fixing a car, brewing coffee, folding laundry. That “egocentric” footage, captured from the perspective of the person performing the task, is one of several methods robotics companies use to build training sets, alongside teleoperation, in which a human remotely drives a robot through the same motion.
The collection approach has real cost advantages. Teleoperation requires expensive hardware, skilled operators, and a robot for every hour of data collected. Paying contributors with smartphones and off-the-shelf sensors scales more cheaply, though the resulting data generally needs more processing before it is usable.
Scale is the selling point. As of early June, Mecka projected it would end 2026 at an annual run rate of $100 million, Gao told Fortune when the previous round was announced. The company has not publicly disclosed its customer list, but robotics developers and AI labs are known to draw on third-party physical data for their models.
Why investors are repricing physical data
The broader shift is hard to miss. LLM data markets, such as they exist, are consolidating around a handful of large vendors, while physical data remains fragmented and underpriced. Robots cannot be trained on web text, and simulation, though improving, still struggles with the friction of messy real-world environments.
The clearest validation of the category came last week, when TechCrunch reported that Mecka’s rival XDOF was nearing a round at a $1.2 billion valuation. Human-data platforms are also migrating toward robotics. Scale AI and Micro1, both built around labor-intensive annotation, have extended into physical-world collection as language-model demand matures.
Whether Mecka can defend its position against a better-funded competitor is the open question. Its advantage is a contributor network and a data pipeline that has already been operating for more than a year; its disadvantage is that capital can replicate both, given time and a sponsor willing to spend.
What to watch
Detailed terms of the Sequoia round, including the exact amount raised and any secondary components, are likely to emerge within weeks. Watch for any reference customers Mecka may name publicly, since a $500 million valuation sits awkwardly next to an undisclosed customer base. And keep an eye on whether the round closes at the reported price, or whether the number shifts as competitors argue for higher marks.
The $500 million figure compares with a prior round whose terms were never fully disclosed. If it holds, Mecka will be valued at roughly five times its projected 2026 run rate, a multiple that assumes the physical-data market widens rather than consolidates around one or two vendors. That assumption is the real bet, and it is not fully settled.
Warning: Attempt to read property "term_id" on false in /www/wwwroot/stockpil.com/wp-content/themes/flex-mag/functions.php on line 998
