NVIDIA’s Nemotron 3 Ultra tops benchmarks with LangChain agent harness

In brief: On 8 July NVIDIA highlighted benchmark-leading results for Nemotron 3 Ultra, its latest open model, saying it delivers leading performance at lower cost than top closed models. The results stem from a collaboration with LangChain, which optimised its Deep Agents harness for the model, aimed at teams building AI agents on open models.

NVIDIA has highlighted benchmark-leading results for Nemotron 3 Ultra, its latest open model, saying it delivers leading performance at lower cost than top closed models. The company announced the milestone on July 8.

The results are tied to a collaboration with LangChain, which optimized its Deep Agents harness specifically for Nemotron 3 Ultra. NVIDIA describes LangChain as the largest and most widely adopted AI agent orchestration platform — the software layer that lets a model plan, call tools and carry out multi-step tasks.

The pairing is aimed at teams building AI agents on open models rather than closed, proprietary systems, where NVIDIA is pitching a lower total cost for comparable capability.

Nemotron 3 Ultra is part of NVIDIA’s broader push to supply not just chips but the models and software stacks that run agentic AI workloads.

Read more: NVIDIA’s Nemotron 3 Ultra tops benchmarks with LangChain agent harness · Apple to launch Apple Intelligence in China in December with iOS 27.3 · OpenAI launches GPT-5.6 with Sol, Terra and Luna models

Sources: NVIDIA Blog.

Fact-checked against the linked sources before publication. Spotted an error? See our Corrections Policy.

Spotted an error? Email corrections@itmagazineme.com with the article link.