Models & Labs
NVIDIA Unveils Cosmos 3 World Foundation Model
Sam Witteveenhigh confidence
Why it matters
- →Cosmos 3's omnimodal capabilities allow for more integrated AI applications.
- →It enhances NVIDIA's position in the AI market by offering advanced multi-modal processing.
- →Developers gain new tools for creating versatile AI solutions.
NVIDIA has introduced Cosmos 3, its latest world foundation model, which is designed to be omnimodal, capable of processing five different types of inputs and generating five types of outputs. This model is part of NVIDIA's push towards enhancing physical AI and open-world reasoning capabilities. Cosmos 3 is expected to provide developers with new opportunities to create more integrated and versatile AI applications. This development underscores NVIDIA's commitment to advancing AI technology and expanding its application across various domains.
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 Nemotron 3 Available on Together AI
Together AI Blog · March 11, 2026 · Related
- 2Nvidia Releases 120B Parameter Nemotron Model
Lev Selector · March 13, 2026 · Background
- 3Together AI Launches NVIDIA Nemotron 3 Nano Omni
Together AI Blog · April 28, 2026 · Related
- 4NVIDIA Launches Nemotron 3 Nano Omni
Matt Wolfe · May 1, 2026 · Related
- 5Google DeepMind Launches Gemini 3.5 Flash Model
Google DeepMind · May 15, 2026 · Background
- 6Fine-Tuning NVIDIA Cosmos Predict 2.5 with LoRA/DoRA
Hugging Face Blog · May 18, 2026 · Related
- 7Google Unveils Gemini Omni and 3.5 Models
Google AI Blog · May 19, 2026 · Background
- 8Google Unveils Gemini 3.5 with Enhanced AI Capabilities
Google AI Blog · May 19, 2026 · Background
What happened next
- 9NVIDIA Launches Nemotron 3 Ultra Model
Ollama Blog · June 4, 2026 · Related
- 10NVIDIA Unveils Nemotron 3 Ultra Model
Sam Witteveen · June 4, 2026 · Related
- 11NVIDIA Showcases AI Advances at SIGGRAPH
NVIDIA Blog · July 20, 2026 · Same story
- 12NVIDIA Releases Cosmos 3 Edge for Robotics
Hugging Face Blog · July 20, 2026 · Same story
- 13NVIDIA Cosmos 3 Advances Open World Models for Physical AI
NVIDIA Blog · August 6, 2026 · Same story
Follow this story
Open the full story →NVIDIA Launches Cosmos 3 for Physical AI
3 developments
- Jun 1 · Hugging Face Blog
NVIDIA Launches Cosmos 3 for Physical AI - Jun 1 · NVIDIA Blog
NVIDIA Unveils Cosmos 3 for Physical AI Development - Jun 1 · Sam Witteveen
NVIDIA Unveils Cosmos 3 World Foundation Model (This article)
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