NVIDIA Nemotron 3 Ultra is being positioned as a lower-cost option for AI agent workflows after LangChain tuned its Deep Agents harness for the model. According to the source, the tuning delivered the highest accuracy among open models on LangChain’s Deep Agents benchmark, while completing more tasks at higher throughput and running at 10x lower inference cost per run than leading closed models.
Measured against LangChain’s Deep Agents benchmark, Nemotron 3 Ultra also achieved business task parity with the highest-scoring closed models. No model retraining was required. The gains came from changes to the environment around the model, including system prompts, tool descriptions and middleware.
LangChain said its agent engineering platform has more than 200 million monthly downloads. The company’s tuned Deep Agents harness for NVIDIA Nemotron 3 Ultra is available directly through LangChain, giving developers access to the profile without retraining the model.
The release also includes NVIDIA NemoClaw for LangChain Deep Agents, described as an open reference blueprint for enterprises building specialized AI systems of models, tools and runtime for their own workflows. It combines LangChain Deep Agents Code, tuned for Nemotron 3 Ultra, with the NVIDIA OpenShell secure runtime for executing agent actions safely.
The source says this matters for enterprises because an open model, an open harness and an open secure runtime let them own the full stack end to end. It also says that becomes more important as agents move from answering questions to taking action inside core systems.
LangChain developers can access Nemotron 3 Ultra on Baseten, Crusoe Cloud, DeepInfra, Fireworks, Nebius and Together AI platforms, which provide a hosted path to the tuned harness in production.
Abridge, Amdocs and Box are embedding specialized agents directly into their platforms, while EY is expanding its NVIDIA implementation capabilities around NVIDIA NemoClaw blueprints for LangChain Deep Agents to help clients customize, evaluate and govern specialized agents across high-value workflows.
Harrison Chase, cofounder and CEO of LangChain, said: “The way to build better agents is to keep improving the system around the model,” and added: “Memory, tool use, evaluation and model behavior compound when teams can tune them together. Our work with NVIDIA shows that enterprises can get strong performance from an open stack while keeping control over the agent systems they are building.”
The source says NVIDIA founder and CEO Jensen Huang recently sat down with Chase to discuss why the last six months have seen a leap in useful AI for enterprises.
NemoClaw for LangChain Deep Agents and the tuned Nemotron 3 Ultra model profile are available now. Developers can pull the tuned Deep Agents harness directly from LangChain, or use the NemoClaw for LangChain Deep Agents blueprint as a starting point for building specialized agents from scratch.
Source: blogs.nvidia.com.
Companies can share verified announcements through Newz9’s international press release submission page.
