Navan (NASDAQ: NAVN) has deployed Nvidia Nemotron 3.5 Super VL to power several of its AI travel agents in production, according to a company press release. The model was trained and deployed within three weeks of Navan receiving early access from Nvidia.

Before deployment, Navan internally evaluated Nemotron 3.5 Super VL against its self-hosted production baseline and other models using consistent prompts, tools, travel scenarios, and criteria. After fine-tuning with Nvidia NeMo AutoModel VLM LoRA training recipe, Nemotron 3.5 Super VL beat or tied competing models on six of seven measures, including lowest latency, based on Navan's internal evaluation.

Navan said the model is running in production at a lower cost compared to its previous baseline. The company described the deployment as part of its open-model strategy, which involves evaluating models for specific travel tasks, adapting them with travel-specific data, and self-hosting them to maintain control over performance, cost, and reliability.

"Building reliable agentic travel systems involves much more than choosing a model, but continuously iterating to improve and solve real-world problems," said Ilan Twig, co-founder and CTO of Navan. "We're grateful to Nvidia for early access to Nemotron 3.5 Super VL, which shows how open models can help us move faster, maintain control, and deliver meaningful value without compromising quality."

Navan said it combines open and closed models to build AI travel agents designed to interpret traveler preferences, enforce company policies, and execute bookings.