Mecka raises $60M Series B to turn human motion into robot training data
What's the deal? Mecka AI has raised a $60 million Series B led by Sequoia CapitalDealroom has a profile for this one. Try Dealroom → to build an internet-scale dataset for training robots. NVIDIA, M12 (Microsoft's venture fund), Qualcomm Ventures, and SamsungDealroom has a profile for this one. Try Dealroom → joined, alongside returning backers Framework VenturesDealroom has a profile for this one. Try Dealroom →, Kindred VenturesDealroom has a profile for this one. Try Dealroom →, and NeoDealroom has a profile for this one. Try Dealroom →.
The round, announced in October 2026, also drew angels Tony XuDealroom has a profile for this one. Try Dealroom →, Frank SlootmanDealroom has a profile for this one. Try Dealroom →, and Milan Kovac. Mecka did not disclose a post-money valuation; earlier discussions pointed to roughly $500 million, a figure not tied to this financing.
What's the endgame? Mecka records egocentric data — footage and motion captured from the first-person view of the person doing a task. It uses body sensors, smartphones, and specialised hardware, paying participants to demonstrate everyday activities in real-world settings rather than labs.
Its in-house lab handles motion tracking and 3D reconstruction, turning flat video into spatial data a machine can learn from. The company has built a dataset called EgoVerse, aimed at developers working on humanoid robots and embodied AI.
By the numbers: The New York and Toronto startup has now disclosed $128 million across three rounds — an $8 million seed in August 2025 and a $60 million Series A in tranches across November 2025 and June 2026. It reported annual run-rate revenue above $100 million as of June 2026 and expects to reach $300 million by year end.
That profile is unusual: a team of 40 to 60 people posting nine-figure run-rate revenue in a sector better known for burning cash than collecting it.
What could go wrong? Tripling run-rate from above $100 million to $300 million in six months demands serious momentum. There's also a technical open question — how well human data transfers to machines, since a human hand and a robot gripper are not the same tool.
The signal: At $60 million, the round lands in the top fifth of deals by size, a sign of the capital chasing the robotics data layer. Mecka is betting that recording people, not robots, is the fastest route to training the next generation of machines.
Read more: cryptobriefing.com
Image credit: Mecka AI
