Models & Labs
NVIDIA Unveils Nemotron 3 Ultra Model
Sam Witteveenhigh confidence
Why it matters
- →The Nemotron 3 Ultra's large parameter size enhances AI model capabilities.
- →Its agentic focus makes it ideal for developing LLM agents.
- →Multi-teacher distillation improves model adaptability and performance.
NVIDIA has launched the Nemotron 3 Ultra, a new AI model with 550 billion parameters, of which 55 billion are active. This model is part of NVIDIA's Nemotron 3 series, which includes the Nano and Super variants. The Nemotron 3 Ultra is designed with a focus on agentic applications, utilizing multi-teacher distillation and post-training for enhanced performance. This development highlights NVIDIA's ongoing efforts to push the boundaries of AI technology, providing a powerful resource for developers working on advanced AI projects.
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 · Same story
- 2Nvidia Releases 120B Parameter Nemotron Model
Lev Selector · March 13, 2026 · Same story
- 3Together AI Launches NVIDIA Nemotron 3 Nano Omni
Together AI Blog · April 28, 2026 · Same story
- 4NVIDIA Launches Nemotron 3 Nano Omni Model
NVIDIA Blog · April 28, 2026 · Same story
- 5NVIDIA Launches Nemotron 3 Nano Omni
Matt Wolfe · May 1, 2026 · Same story
- 6NVIDIA Launches Cosmos 3 for Physical AI
Hugging Face Blog · June 1, 2026 · Related
- 7NVIDIA Unveils Cosmos 3 World Foundation Model
Sam Witteveen · June 1, 2026 · Related
- 8Nvidia Unveils AI Agent-Centric Tech at COMPUTEX
The Rundown AI · June 2, 2026 · Same story
What happened next
- 9NVIDIA Nemotron 3 Ultra Boosts AI Agent Performance
NVIDIA Blog · July 8, 2026 · Same story
- 10NVIDIA's Nemotron Enhances AI with Open Synthetic Data
Hugging Face Blog · July 8, 2026 · Same story
- 11NVIDIA Nemotron 3.5 Lightning Launches for Local AI Agents
Ollama Blog · August 11, 2026 · Same story
- 12NVIDIA Unveils Nemotron Lightning Model
Sam Witteveen · August 11, 2026 · Same story
- 13NVIDIA Unveils Nemotron 3.5 Lightning
Lev Selector · August 14, 2026 · Same story
Follow this story
Open the full story →NVIDIA Launches Nemotron 3 Ultra Model
3 developments
- Jun 4 · Ollama Blog
NVIDIA Launches Nemotron 3 Ultra Model - Jun 4 · Sam Witteveen
NVIDIA Unveils Nemotron 3 Ultra Model (This article) - Jun 5 · Matt Wolfe
NVIDIA Unveils Nemotron 3 Ultra for AI Agents
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