NVIDIA has released Cosmos 3 Edge, a new 4-billion-parameter open-source AI model designed for on-device operation in robots and autonomous vehicles. This model can process various inputs including text, images, video, and motion trajectories, and output actions, text, images, or video. It is capable of real-time reasoning, generating up to 32 robot actions per inference and running at 15 Hz on edge devices like NVIDIA's Jetson Thor and consumer-grade GeForce RTX graphics cards, enabling local processing without an internet connection. AI
IMPACT Enables real-time reasoning and action generation for robots and autonomous vehicles on edge devices.
RANK_REASON Frontier-lab model release with system card. [lever_c_demoted from frontier_release: ic=2 ai=1.0]
Read on Mastodon — fosstodon.org →
AI-generated summary · Google Gemini · from 2 sources. How we write summaries →