Key facts
- Nvidia is developing Nemotron 4, a new family of open-source AI models.
- The largest Nemotron 4 model is expected to have at least 1 trillion parameters.
- Nvidia has released Nemotron 3.5 Lightning, a 30-billion-parameter mixture-of-experts model for agentic workflows.
- Nemotron 3.5 Lightning delivers up to 4x faster output speed compared to similar models.
- Nvidia also launched NeMo Switchyard, an open-source library for directing AI requests to appropriate models.
Nvidia is developing Nemotron 4, a new family of open-source AI models, with the largest version potentially exceeding 1 trillion parameters. This initiative aims to compete with global open-source AI offerings. The company also announced the release of Nemotron 3.5 Lightning, a 30-billion-parameter mixture-of-experts model designed for long-running agentic AI workloads. This model offers up to four times the output speed of similar-sized models and enables 30% faster agentic task completion. Nvidia further introduced NeMo Switchyard, an open-source library for intelligently routing AI requests to the most suitable model. CEO Jensen Huang has expressed support for open-weight AI models, viewing them as crucial for American AI leadership. The release follows Meta's recent launch of its open-weight Muse Glimmer AI Model.