en.Wedoany.com Reported - Glean announced support for the NVIDIA Nemotron 3 Ultra model, adding this open-source model option to its platform to help customers achieve cost-effective agentic applications in daily enterprise workflows.

Nemotron 3 Ultra, as an open-source model, reportedly achieves 91% of the completeness of frontier large language models at open-source costs. With this model now available on the Glean platform, enterprises gain greater flexibility when deploying AI across business operations. Glean does not mandate binding all tasks to a single model family; instead, it helps organizations select the most suitable model for specific tasks and orchestrate them within an enterprise platform focused on security and context awareness.

Emrecan Dogan, Chief Product Officer at Glean, stated that enterprises are moving beyond the "one model for all tasks" mindset and seek the ability to match the right model with the right task, bringing AI into daily work cost-effectively. He emphasized that support for NVIDIA Nemotron 3 Ultra aligns with this trend, providing customers with a powerful option as they scale AI applications.

Kari Briski, Vice President of Generative AI at NVIDIA, noted that Glean is bringing Nemotron 3 Ultra into enterprise AI workflows, where model selection, cost, and performance are critical requirements. The two companies are jointly helping enterprises deploy open-source models to support daily operations at scale.

This announcement reflects Glean's long-standing model-agnostic platform strategy, where enterprises should build freely within a model ecosystem rather than relying on a single vendor. With over 30 models, including Nemotron 3 Ultra, Glean customers can leverage the latest open-source and proprietary advancements to achieve stronger performance, lower costs, and avoid vendor lock-in amid rapid AI iteration.

Glean's collaboration with NVIDIA on the Nemotron model family has a track record. Previously released Glean Waldo, an agentic search model, was post-trained on NVIDIA Nemotron 3 Nano, achieving a 50% reduction in latency and a 25% reduction in tokens. Waldo offloads search tasks previously handled by frontier models, allowing them to focus on scenarios requiring higher reasoning and response capabilities. This reflects Glean's token economy approach: multiple models working together to deliver frontier-level intelligence with fewer tokens.

Glean provides the context and intelligence layer for enterprise AI. Its assistant product leverages the Glean enterprise knowledge graph to offer employees an AI assistant based on company data; its agentic capabilities enable teams to create, use, and manage AI agents through natural language. Relying on its search and agent engine, Glean helps organizations automate work at scale while enforcing permission controls, maintaining traceability, governance, and security. Through native integrations, MCP servers, model selection, and customizable APIs, Glean offers enterprises an open, scalable pathway to deploy complex AI ecosystems on a single horizontal platform.

This article is compiled by Wedoany. All AI citations must indicate the source as "Wedoany". If there is any infringement or other issues, please notify us promptly, and we will modify or delete it accordingly. Email: news@wedoany.com