Models & Labs
NVIDIA Releases Cosmos 3 Edge for Robotics
Hugging Face Bloghigh confidence
Why it matters
- →Cosmos 3 Edge enables real-time AI reasoning and action generation on edge devices.
- →It integrates multimodal data processing, enhancing robotics and vision AI capabilities.
- →The model supports efficient deployment in memory-constrained environments, broadening AI application scope.
NVIDIA has launched Cosmos 3 Edge, a 4-billion-parameter model designed for edge devices, available on Hugging Face. This model is tailored for real-time reasoning and action generation in robotics and vision AI, operating efficiently on NVIDIA's edge computing platforms. Cosmos 3 Edge integrates two transformer towers to process multimodal data, enabling robots to understand and interact with their environment. This release enhances the capability of AI systems to perform complex tasks directly on-device, improving their adaptability and efficiency in various settings.
Read originalThe story around this
Earlier coverage that leads up to this article, and what followed. Lines connect each piece to the closest one after it, converging here.
How we got here
- 1NVIDIA Launches Nemotron 3 Nano Omni
Matt Wolfe · May 1, 2026 · Related
- 2NVIDIA Launches Cosmos 3 for Physical AI
Hugging Face Blog · June 1, 2026 · Same story
- 3NVIDIA Unveils Cosmos 3 for Physical AI Development
NVIDIA Blog · June 1, 2026 · Same story
- 4NVIDIA Unveils Cosmos 3 World Foundation Model
Sam Witteveen · June 1, 2026 · Same story
- 5Nvidia Unveils AI Agent-Centric Tech at COMPUTEX
The Rundown AI · June 2, 2026 · Related
- 6Nvidia Unveils Blueprint for Advanced Humanoid Robots
WIRED AI · June 3, 2026 · Related
- 7NVIDIA Launches Nemotron 3 Ultra Model
Ollama Blog · June 4, 2026 · Related
- 8NVIDIA Unveils Nemotron 3 Ultra Model
Sam Witteveen · June 4, 2026 · Related
What happened next
- 9Google DeepMind Unveils Gemini Robotics 2
WIRED AI · July 30, 2026 · Related
- 10NVIDIA Cosmos 3 Advances Open World Models for Physical AI
NVIDIA Blog · August 6, 2026 · Same story
- 11NVIDIA Jetson Orin Nano 2 targets edge AI for robots
AI News · August 26, 2026 · Related
- 12Nvidia releases open Nemotron models for agents and PCs
TechCrunch AI · October 19, 2026 · Related
- 13NVIDIA Cosmos 3 Open Models for World Simulation
Hugging Face Blog · November 5, 2026 · Same story
Follow this story
Open the full story →Open Models Propel AI Research at ICML 2026
3 developments
- Jul 6 · NVIDIA Blog
Open Models Propel AI Research at ICML 2026 - Jul 20 · NVIDIA Blog
NVIDIA Showcases AI Advances at SIGGRAPH↳ NVIDIA unveiled the Cosmos 3 Edge model for real-time on-device processing at SIGGRAPH. - Jul 20 · Hugging Face Blog
NVIDIA Releases Cosmos 3 Edge for Robotics (This article)↳ NVIDIA released Cosmos 3 Edge, a 4-billion-parameter model for robotics that enables real-time reasoning on memory-constrained edge devices.
More from Hugging Face Blog
© Hugging Face BlogModels & Labscoding
IBM releases LLMON markup to fix prompt injection
IBM is tackling the brittleness of small language models by introducing LLMON, a structured markup language that explicitly separates instructions from data. By using reserved special tokens instead of ambiguous characters like JSON brackets, these new Granite 4.1 preview models drastically reduce indirect prompt injection attacks while maintaining standard performance. This approach forces the model to bind specific constraints to specific data payloads, eliminating the confusion that often leads to hallucinations or security breaches in long contexts.
Hugging Face Blog
© Hugging Face BlogModels & Labsother
Hugging Face releases TRELLIS.2 fine-tuning guide
Hugging Face finally provides a concrete path to adapt the powerful TRELLIS.2 model for specific domains like footwear or furniture, solving the 'generalist' problem in 3D generation. By targeting only the structure and shape DiTs with LoRA adapters while freezing the texture flow, they prevent catastrophic forgetting of material priors—a common pitfall when fine-tuning large generative models. This approach allows practitioners to achieve high geometric accuracy on small datasets (188 samples) without retraining the entire pipeline. It transforms TRELLIS.2 from a demo toy into a usable tool for production-grade 3D asset creation.
Hugging Face Blog
© Hugging Face BlogOpen Sourceagents
Tether AI launches BotPsy, a sub-billion agentic model
Tether AI Research has released BotPsy, an 0.8B-parameter model built on Qwen3.5 that specializes in tool-use and agentic workflows for on-device deployment. It achieves state-of-the-art performance in its size class, outperforming larger models like Qwen3.5-2B-Think on benchmarks such as BFCL v4 and ACEBench while maintaining strict privacy by keeping data local. The model is designed to handle routine tasks like calendar management and web search without relying on expensive cloud APIs, offering a viable path for private, self-hosted personal assistants. With open weights under Apache 2.0 and GGUF support, it provides immediate utility for developers building lightweight, autonomous agents.
Hugging Face Blog
More in Models & Labs
Models & Labsother
llama.cpp b12340 adds CUDA 13 and ROCm 10.0
This release quietly closes the hardware gap for local inference by adding default support for CUDA 13 and ROCm 10.0 across Linux and Windows. Users with newer NVIDIA GPUs or AMD accelerators no longer need to compile from source to get performance, while Apple Silicon builds now ship KleidiAI disabled by default, shifting the focus back to standard ARM optimizations. The inclusion of Snapdragon support for Linux and Android marks a significant step toward on-device AI beyond the x86/ARM desktop ecosystem. It is less about new features and more about ensuring llama.cpp remains the universal runtime for every available accelerator.
llama.cpp Releases
Models & Labscoding
llama.cpp b12342 adds few-row MMQ for speculative decoding
Speculative decoding just got a significant speed boost on NVIDIA GPUs. This release introduces a specialized kernel for 'few-row' matrix multiplications, targeting the small batch sizes typical of verification passes in speculative decoding workflows. By optimizing these specific operations for quantized formats like Q8_0 and IQ4 variants, llama.cpp reduces latency where it matters most during token generation. The change is CUDA-only for now but leverages hardware capabilities from Ampere onwards, making local inference faster without requiring new model architectures or complex configuration changes.
llama.cpp Releases
© TechCrunch AIModels & Labsmodels
Google deploys Gemini 4 Hydrogen for faster Search
Google is quietly shifting its search infrastructure by replacing older fast models with Gemini 4 Hydrogen, a tier designed specifically for speed and cost-efficiency rather than raw reasoning power. This model completes tasks 2.5 times faster than its predecessor while beating the more powerful Argon on specific intelligence benchmarks, proving that specialized efficiency can outperform brute force in high-volume workflows. By integrating this into Search's fan-out technique, Google now breaks down complex queries into subqueries with greater precision and speed, delivering richer visual responses without the latency of heavier models. This marks a strategic pivot toward agentic execution layers within consumer search, making AI Mode feel snappier for everyday multi-part questions.
TechCrunch AI