TwelveLabs Pegasus 1.6 Brings Egocentric Video Understanding to Physical AI

TwelveLabs Inc. says the physical world is digitalizing quickly, and AI teams now gather huge volumes of complex video. Turning that raw footage into models that can act remains difficult.

The company released Pegasus 1.6, a new model that adds capabilities for understanding and navigating complex real-world environments.

Jae Lee, co-founder and CEO of TwelveLabs, said the company's mission "has always been to help machines understand how the world works through video."

He called physical AI "the next expression of that mission," noting that much of what people know about physical work — such as adjusting a grip or recovering after a slip — has never been captured in a form a machine can learn from.

Lee said the new release turns such footage into "structured, reviewable knowledge," so robotics and physical AI teams can train on real human experience instead of starting from scratch.

Founded in 2021 and based in Seoul, TwelveLabs says it has built a full-stack video intelligence platform using its Marengo and Pegasus models.

Designed to see and understand video the way humans do but far faster, the system improves over time, and developers can use it to access and act on their video content.

The latest Pegasus release marks TwelveLabs' expansion into physical AI, generating rich insights from real-world perspectives.

The company claims it solves workflow-specific challenges so robots, drones and autonomous vehicles can perceive, reason and act safely in the physical world.