Mecka AI Nears $500M Valuation in Sequoia Deal Amid Physical Data Rush

Mecka AI is closing in on a new funding round led by Sequoia Capital that values the two-year-old human motion data startup at approximately $500 million, according to sources familiar with the transaction. The capital injection arrives just months after the company secured a $60 million total investment led by Framework Ventures, reflecting accelerating investor competition for physical-world training data needed by general-purpose robotics labs.

Derived from the fictional giant robot concept known as mecha, the startup was established in 2024 by co-founders Josh Gao, Mogen Cheng, Jason Chong, and Duy Nguyen. Operating primarily out of Toronto with a team of between 40 and 60 employees, the company bypasses traditional synthetic simulation and clunky remote-control teleoperation by deploying custom body sensors and iPhones to record everyday, unscripted human activities.

At the core of the startup's commercial offering is Egoverse, a proprietary dataset capturing first-person human interactions and physical maneuvers such as operating kitchen equipment or performing vehicle maintenance. Robotics developers and frontier AI model teams license this repository to train robotic policies for complex real-world environments. Co-founder Josh Gao disclosed to Fortune in early June that the enterprise was projecting an annual recurring revenue run rate of $100 million by the conclusion of 2026, though specific underlying client agreements have remained private.

Market demand for physical-world data pipelines has surged as autonomous hardware developers seek alternatives to lab-concocted testing. Mecka's rapid scaling mirrors the trajectory of human-data platforms traditionally focused on large language models, positioning physical AI data collection as a foundational bottleneck for humanoid development. Other market participants are pursuing parallel valuation milestones, including physical data startup XDOF, which TechCrunch reported was approaching a $1.2 billion valuation in its own financing talks.

Before entering advanced talks with Sequoia Capital, Mecka established a multi-stage funding trail starting with a $25 million Series A round in November 2025. This was followed by a $35 million extension announced in June 2026, both helmed by Framework Ventures with additional backing from Menlo Ventures, SV Angel, Kindred Ventures, and angel investor Ted Xiao, bringing cumulative capital raised to approximately $68 million.

Alongside its June 2026 funding extension, the company completed the acquisition of Docula to bolster its internal video-understanding lab. This division processes raw egocentric footage into structured, training-ready formats required by machine learning pipelines, streamlining the conversion of messy human physical behavior into actionable model weights.

Framework Ventures, the lead backer across Mecka's Series A rounds, maintains deep roots in cryptocurrency and decentralized blockchain infrastructure. This institutional crossover highlights an evolving trend where web3-aligned venture funds target physical AI infrastructure, driven by emerging commercial questions surrounding data ownership, cryptographic provenance tracking, and permissioned access to proprietary training corpuses.

While hardware and robotics developers grapple with the logistical challenges of data security and verification, investors are betting that decentralized provenance models could eventually intersect with physical data collection. As AI labs ingest millions of hours of human motion footage, establishing clear rights management and verifiable data trails is becoming an urgent priority across the sector.

The broader physical AI sector encompasses competing approaches to robot deployment and autonomy, highlighted by rival firms entering the commercial market. Industrial automation startup Maven Robotics recently emerged from stealth with $100 million in backing from RoboStrategy, LocalGlobe, Vine Ventures, and XTX Ventures, deploying wheeled systems for warehouse palletizing rather than focusing purely on humanoid models.

humanoid robotics developer Agility prepared for a $2.5 billion public debut via a special purpose acquisition company deal. As startups across the ecosystem vie for enterprise contracts and capital, the primary constraint remains the acquisition of high-fidelity physical training data capable of bridging the gap between simulated environments and messy industrial reality.