LangChain has introduced the NemoClaw for LangChain Deep Agents Blueprint, developed in collaboration with NVIDIA to help enterprises build, evaluate, deploy, and optimize open AI agent systems. The blueprint combines LangChain's Deep Agents framework with NVIDIA Nemotron 3 Ultra and NVIDIA OpenShell to provide organizations with a secure, customizable, and cost-efficient foundation for deploying production-ready AI agents.
LangChain announced the launch of the NemoClaw for LangChain Deep Agents Blueprint, providing enterprises with an open architecture for developing AI agents that can be customized, deployed securely, and continuously optimized.
As organizations increasingly move AI agents into production, enterprise teams require greater ownership over workflows, memory, model tuning, and operational governance. The new blueprint addresses these requirements by allowing businesses to retain control over proprietary AI assets while improving performance and reducing operational costs.
The NemoClaw blueprint integrates three core technologies into a unified enterprise AI development stack.
NVIDIA Nemotron 3 Ultra serves as the open-weight foundation model that organizations can customize for domain-specific use cases.
LangChain Deep Agents provides the orchestration layer, supporting long-running AI agents with capabilities including planning, memory management, tool usage, reasoning, and task execution.
NVIDIA OpenShell functions as the runtime layer, enabling secure deployment while governing how AI agents interact with enterprise systems, applications, and data.
Together, these components allow enterprises to deploy AI agents that can be measured, governed, optimized, and scaled within production environments.
According to LangChain's internal evaluation benchmarks, NVIDIA Nemotron 3 Ultra running with LangChain Deep Agents achieved an aggregate score of 0.86 while costing $4.48 for inference.
The company noted that the closest competing model delivered similar performance at $43.48, representing approximately 10 times higher inference costs.
LangChain attributes the improvements to customized tuning across model behavior, tool usage, context management, and workflow execution rather than relying solely on model performance.
"The way to build better agents is to keep improving the system around the model," said Harrison Chase, Co-founder and CEO of LangChain. "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're building."
The blueprint is designed to help organizations create specialized AI agents that can evolve over time through continuous evaluation, workflow optimization, and governance.
"Super agents have arrived," said Jensen Huang, Founder and CEO of NVIDIA. "With an open model like NVIDIA Nemotron, a LangChain harness, the NVIDIA OpenShell runtime, and a company's own data, every enterprise can build custom agents that understand its business, use its tools, and turn knowledge into action. The future of AI won't be one-size-fits-all — companies will use AI cloud services and build their own AI, shaped by their proprietary data, know-how, and workflows, and run it safely and securely wherever they operate."
The announcement is supported by ecosystem partners including EY, Baseten, Fireworks AI, Nebius, Crusoe, DeepInfra, and Together AI, which will help enterprises deploy and scale AI agent systems across production environments.
The NemoClaw for LangChain Deep Agents Blueprint is available immediately for enterprises looking to evaluate and deploy open AI agent architectures for their own workloads.
LangChain powers the full agent development lifecycle building, testing, deploying, and monitoring — so AI teams can improve their agents systematically. LangSmith Engine accelerates this cycle, automatically surfacing and fixing issues to improve agents over time. LangSmith is neutral by design, so teams can customize their own stack to optimize on cost and performance as the landscape evolves. More than 7,000 customers, including NVIDIA, Bridgewater, LinkedIn, Workday, Harvey, and Rippling trust LangSmith to build and manage their agents.
