Nvidia Unveils 30B-Parameter Nemotron 3.5 Lightning AI Model

Nvidia has released Nemotron 3.5 Lightning, a 30-billion-parameter mixture-of-experts model for long-running agentic AI workloads, and NeMo Switchyard, an open-source library for smart routing inside popular agent tools. The new model has 3 billion active parameters and delivers up to 4x faster output speed and 30% faster agentic task completion compared to other models in its class.

Nemotron 3.5 Lightning is licensed under OpenMDW-1.1, which permits commercial use. Nvidia has published the training data and recipes for the model alongside the weights. In benchmark tests, Nemotron 3.5 Lightning achieved 81.94 on MMLU Pro, 75.44 on GPQA Diamond, and 51.56 on SWE-bench Verified.

The release of Nemotron 3.5 Lightning and NeMo Switchyard is part of Nvidia's efforts to expand its Nemotron model family. NeMo Switchyard can route tasks to the most suitable models, reducing costs by up to 74% compared to using a frontier model alone. This allows enterprises to build a router based on their specific needs and deploy it without requiring developers to rewrite their applications.

## What's Next Nvidia is also developing Nemotron 4, a new AI model family with at least 1 trillion parameters. The largest Nemotron 4 model is expected to be ready as early as late fall, but Nvidia has not set a release date. This new model family is aimed at rivaling top open-source models globally.

## Why it Matters The release of Nemotron 3.5 Lightning and NeMo Switchyard is significant as AI shifts from chatbots to autonomous agents. Open models are serving market demands for full control over where AI runs and how it's deployed and evolves. Nvidia's commitment to continually improving open models for greater accuracy and speed is reflected in the release of Nemotron 3.5 Lightning and NeMo Switchyard. These models and libraries deliver greater control over how AI is deployed, where it runs, and how efficiently it operates — across PCs, workstations, data centers, and the cloud.

The development of Nemotron 4 is also noteworthy, as it represents Nvidia's efforts to stay ahead in the AI market. The company's focus on open-source models and collaboration with other companies to develop and share tools for AI safety and cybersecurity underscores its commitment to responsible AI development. As the AI landscape continues to evolve, Nvidia's releases and developments are likely to have a significant impact on the industry.