MindOn claims 142 garments an hour in uncut Mind-1 folding demo

The nearly hour-long video shows a dual-arm station folding at about 25 seconds per garment — with a person still handing each item to the robot.

Shenzhen startup MindOn has released a nearly hour-long, uncut video of its Mind-1 system folding clothes, claiming throughput of 142 garments per hour — roughly 25 seconds per item.

The footage shows a fixed dual-arm station lifting, spreading and folding shirts on a work surface, with an on-screen counter tracking completed items. The video is labeled as running at real-time speed with fully autonomous operation.

One caveat is visible throughout: a person places a new garment on the table after each completed fold. The folding itself is automated; feeding the next item is still a human job.

MindOn calls the result the industry's fastest robotic clothes folding. That superlative is the company's own claim — no side-by-side benchmark against other systems has been published, and garment choice, starting conditions and fold quality all affect the number.

The video is a follow-up to MindOn's September 30 debut of Mind-1, which showed parcel handling, shirt folding and laundry collection at what the company described as human speed. The throughline of both releases is software, not hardware: Mind-1 is a physical AI model designed to run across different robot bodies, from Unitree G1 humanoids to fixed dual-arm stations.

The longer format matters. Brief highlight reels hide interventions and cherry-picking; an hour of continuous footage gives skeptics far more to inspect — even if the hardest part of the laundry workflow, getting each garment to the station, remains conspicuously human.