nvidia-cosmos-61dc1f4c·4 events·first seen Aliases: NVIDIA Cosmos, NVIDIA Cosmos 3, Nvidia Cosmos 3
A Latent Space AI news digest covers three NVIDIA announcements: Cosmos 3 (a world model/simulation platform), Nemotron 3 Ultra (a large language model), and RTX Spark (likely a new hardware or inference product). The piece frames these as a significant win for Jensen Huang and NVIDIA's AI portfolio. Coverage is commentary-tier aggregation rather than primary technical reporting.
Anthropic released Claude Opus 4.8 with improvements in coding, reasoning, agentic tasks, and notably better uncertainty flagging—approximately four times less likely than Opus 4.7 to let code flaws pass uncommented. Alongside the model, Anthropic introduced dynamic workflows in Claude Code enabling tens to hundreds of parallel subagents for large-scale engineering tasks, an effort-control slider, and a 3x price cut on fast mode. Anthropic also previewed Mythos-class models, positioned above Opus in capability, currently available to a limited set of organizations for cybersecurity work pending broader safety clearance. The same digest covers MiniMax M3 (open-weights, ~60% SWE-Bench Pro), Nvidia's RTX Spark superchip, Cosmos 3 world model, and a GR00T/Unitree robotics partnership.
NVIDIA has released Cosmos 3, described as the first open omni-model targeting physical AI reasoning and action. The model is hosted and announced via Hugging Face, positioning it as an open-weights offering for robotics and embodied AI applications. The announcement highlights multimodal capabilities oriented toward physical world understanding and agent-level action.
NVIDIA has released Cosmos Reason 2, a model designed to bring advanced reasoning capabilities to physical AI applications. The announcement appears on the Hugging Face blog, indicating the model is likely available or accessible through the platform. This represents a continuation of NVIDIA's Cosmos model family targeting robotics and physical world understanding.