NVIDIA and ServiceNow have officially launched Apriel Nemotron 15B, a compact and efficient open-source large language model (LLM), as part of a joint initiative to accelerate Artificial Intelligence adoption across enterprises. The unveiling took place during ServiceNow´s Knowledge 2025 event, where NVIDIA CEO Jensen Huang and ServiceNow CEO Bill McDermott presented the model and discussed its potential applications in streamlining IT, HR, and customer service workflows with real-time responsiveness.
The Apriel Nemotron 15B model was developed using NVIDIA´s NeMo framework, incorporating the open NVIDIA Llama Nemotron Post-Training Dataset and data specific to enterprise domains from ServiceNow. Training occurred on the NVIDIA DGX Cloud, leveraging Amazon Web Services infrastructure to ensure scalability and robust performance. These technological foundations aim to equip enterprise Artificial Intelligence agents with advanced reasoning capabilities, making them suitable for complex and evolving business tasks.
This announcement follows the earlier release of the NVIDIA Llama Nemotron Ultra model in April, a leading open-source model recognized for its high-level reasoning skills in areas such as scientific research, coding, and advanced mathematics. ServiceNow utilized the same open dataset used in the Ultra model to create Apriel Nemotron 15B, with a focus on cost-efficiency and practical deployment across industry-specific functions. The new model further strengthens the partnership between NVIDIA and ServiceNow, underscoring a commitment to developing accessible Artificial Intelligence technologies designed to scale and enhance organizational processes globally.
