Mecka AI just raised a $60M Series B led by Sequoia.

NVIDIA, Microsoft M12, Qualcomm, and Samsung also participated, along with angels including Tony Xu, Frank Slootman, and former Tesla Optimus lead Milan Kovac. Total funding now exceeds $120 M.

Founded by Josh Gao and co-founders Jason Chong, Mogen Cheng, and Duy Nguyen, none of whom came from traditional robotics labs, the company is building the data and deployment layer for physical AI.

Their core thesis is that the physical world was never recorded at scale. Motion, contact, force, and geometry cannot be scraped from the internet.

Mecka pays people to perform real tasks such as making coffee or doing repairs while wearing custom multi-sensor hardware and phones.

It then converts that egocentric footage into training data using in-house vision models, which Mecka says achieve sub-centimeter hand-pose accuracy on in-the-wild data.

The company has also run a human-to-robot transfer study called EgoVerse with Georgia Tech, Stanford, UC San Diego, ETH Zürich, MIT, and Meta.

The company claims it surpassed $100M in run-rate revenue by June 2026 and is projecting a $300M run rate by year-end, while already supplying frontier robotics labs and multiple Mag 7 companies.

The new capital will scale data collection infrastructure, deepen the research lab, and expand commercial robot deployments.

Physical data remains a core bottleneck for embodied AI. This is one of the more aggressive efforts to industrialize it.