This video shows you how to create a data analyst AI agent using Llama Nemotron Super 1.5 49B, a NAS-optimized, open-weight reasoning model. It tops the Artificial Analysis Intelligence Index in the ~70B range—while fitting on a single NVIDIA H100 or NVIDIA H200 GPU for higher throughput and lower cost.

00:00 — Introduction.
00:39 — What it is.
01:24 — Accuracy.
1:53 — Performance.
02:20 — Open & transparent.
02:59 — Data analysis agent demo.
04:46 — Recap.

Open weights + open dataset on Hugging Face https://huggingface.co/nvidia/Llama-3… from your browser and create API key https://build.nvidia.com/nvidia/llama… Demo repository https://github.com/NVIDIA/GenerativeA… NVIDAI Developer Blog https://developer.nvidia.com/blog/bui… Learn about NVIDIA Nemotron models https://www.nvidia.com/en-us/ai-data–… #LlamaNemotronSuper #NVIDIAAI #OpenSourceAI #GenerativeAI #MachineLearning #DataAnalysis #AIAgents #HuggingFace #GPUComputing #PythonCoding #Pandas #Matplotlib #LLMs #TechDemo.

Try from your browser and create API key https://build.nvidia.com/nvidia/llama…

Demo repository https://github.com/NVIDIA/GenerativeA…