NVIDIA Nemotron 3 Nano 30B MoE model is now available in Amazon SageMaker JumpStart

This article announces the general availability of NVIDIA Nemotron 3 Nano 30B model in Amazon SageMaker JumpStart, a managed deployment service for generative AI applications.

• Nemotron 3 Nano is a small language mixture-of-experts model with 30B parameters and 3B active parameters
• Hybrid Transformer-Mamba architecture with support for token budget optimization
• Excels in coding, reasoning, math, and scientific tasks with leading benchmark performance
• 1 million token context window for extended reasoning capabilities
• Fully open-source with customizable weights, datasets, and recipes
• Managed deployment through SageMaker JumpStart eliminates infrastructure complexity
• Accessible via AWS CLI, SageMaker SDK, and Boto3 for easy integration

Nemotron 3 Nano is now available in SageMaker JumpStart, enabling developers to deploy efficient, high-performance AI models without managing deployment infrastructure.