NVIDIA Nemotron Achieves Benchmark-Leading Performance With LangChain Deep Agents Harness

NVIDIA Nemotron 3 Ultra is offering leading performance at lower cost than top closed models with the largest and most widely adopted AI agent orchestration platform. LangChain tuned its Deep Agents harness for NVIDIA Nemotron 3 Ultra, achieving the highest accuracy among open models, while completing more tasks at higher throughput and running at 10x [...] NVIDIA Nemotron 3 Ultra is offering leading performance at lower cost than top closed models with the largest and most widely adopted AI agent orchestration platform.

LangChain tuned its Deep Agents harness for NVIDIA Nemotron 3 Ultra, achieving the highest accuracy among open models, 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.

Every gain came from engineering the environment around the model, not the model itself. At a tenth of the cost, teams harnessing NVIDIA Nemotron 3 Ultra can run evaluations continuously, experiment faster and build specialized agents across more of their business. LangChain's agent engineering platform has more than 200 million monthly downloads.

By tuning its Deep Agents harness specifically for NVIDIA Nemotron 3 Ultra, it allows for high-performing agents that complete more tasks, run faster and give enterprises a fully open stack they can customize, own and run anywhere. "The way to build better agents is to keep improving the system around the model," said Harrison Chase, cofounder 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 are building." Abridge, Amdocs and Box are embedding specialized agents directly into their platforms and global systems integrator EY is expanding its NVIDIA implementation capabilities around NVIDIA NemoClaw blueprints for LangChain Deep Agents, helping clients customize, evaluate and govern specialized agents across high-value workflows. 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.

Harness Engineering, Not Fine-Tuning LangChain's team ran Nemotron 3 Ultra against its public Deep Agents benchmark suite, then analyzed the deep agent's execution traces to find exactly where it lost points. Instead of retraining the model, the team tuned the harness around it - adjusting system prompts, tool descriptions and middleware. Every developer using LangChain Deep Agents with Nemotron 3 Ultra can put this to work today - the tuned profile is available directly through LangChain.

An Open Stack Built to Own NVIDIA NemoClaw for LangChain Deep Agents is the open reference blueprint that packages this work for enterprises building their own specialized AI - systems of models , tools and runtime - tuned 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. An open model, an open harness and an open secure runtime means enterprises own the full stack, end to end.

They can customize it around the expertise that sets their business apart, keep improving it and run it anywhere - their own infrastructure, their own cloud, their own governance. That distinction matters more as agents take on higher-stakes work. The shift from AI assistants that answer questions to agents that take action inside core systems changes what businesses get from their AI.