Mecka AI raises $60M for human motion data to train robots

Mecka AI has closed a $60 million Series B round led by Sequoia, with participation from Nvidia and Microsoft's venture fund M12, TechCrunch reports. The startup was previously reported to be nearing the round at a $500 million valuation.

How it works

Founded in 2024, Mecka AI collects human motion data to train robots, including humanoid models. It pays individuals to record themselves performing everyday tasks — making coffee, fixing cars — while wearing body sensors and using smartphones. The idea is to do for robotics what data-labeling firms like Scale AI did for large language models: supply human-generated data that AI systems learn from.

The bigger picture

Other startups are chasing the same market. XDOF, which also collects real-world data for robot training, was in talks for a Series B round at a $1.2 billion valuation, per TechCrunch's earlier reporting. Meanwhile, LLM-focused data platforms like Scale AI and Micro1 are expanding into robotics.

This isn't a product you'll buy today. But it's a bet that robots will learn from watching people, and that data is the scarce resource.

• human motion data
• humanoid robots
• Mecka AI
• Microsoft M12
• Nvidia
• Sequoia