Deployment Guides#

Deployment guides, fine-tuning recipes, and agentic usage examples for Nemotron models. Each card links to its directory in the Nemotron GitHub repository.

Nemotron 3 Super

Notebooks for deploying the 120B/12B-active hybrid Mamba-Transformer MoE model with vLLM, SGLang, and TensorRT-LLM.

https://github.com/NVIDIA-NeMo/nemotron/tree/main/usage-cookbook/Nemotron-3-Super
Nemotron 3 Super — LoRA Text2SQL

Supervised fine-tuning with LoRA for Text2SQL using the BIRD SQL benchmark. Includes recipes for both NeMo AutoModel and Megatron Bridge.

https://github.com/NVIDIA-NeMo/nemotron/tree/main/usage-cookbook/Nemotron-3-Super/lora-text2sql
Nemotron 3 Super on DGX Spark

Deploy on a single DGX Spark with 128 GB unified memory using vLLM (nightly) and TensorRT-LLM, including NVFP4 quantization and MTP speculative decoding.

https://github.com/NVIDIA-NeMo/nemotron/tree/main/usage-cookbook/Nemotron-3-Super/SparkDeploymentGuide
Nemotron 3 Ultra Base

550B total / 55B active parameter base model checkpoint announced at GTC 2026. A starting point for custom fine-tuning and RL post-training pipelines — not yet instruction-tuned.

https://github.com/NVIDIA-NeMo/nemotron/tree/main/usage-cookbook/Nemotron-3-Ultra-Base
Nemotron 3 Super on GRPO/DAPO RL Training

Full-weight RL training from a base model using the GRPO/DAPO algorithm to reproduce emergent math reasoning. Requires 5× GB200 or 3× B200 nodes.

https://github.com/NVIDIA-NeMo/nemotron/tree/main/usage-cookbook/Nemotron-3-Super/grpo-dapo
Nemotron 3 Super on Agentic Coding

Use Nemotron 3 Super with OpenCode, OpenClaw, Kilo Code CLI, and OpenHands via OpenRouter and build.nvidia.com.

https://github.com/NVIDIA-NeMo/nemotron/tree/main/usage-cookbook/Nemotron-3-Super/OpenScaffoldingResources
Nemotron Nano 2 VL

Notebooks for the 12B multimodal model that unifies visual and textual understanding. Covers NIM inference via build.nvidia.com and local Hugging Face deployment.

https://github.com/NVIDIA-NeMo/nemotron/tree/main/usage-cookbook/Nemotron-Nano2-VL
Nemotron Parse v1.1

Notebook for the document-parsing VLM that converts PDFs and unstructured documents into structured JSON, LaTeX, and Markdown. Available via NIM at build.nvidia.com.

https://github.com/NVIDIA-NeMo/nemotron/tree/main/usage-cookbook/Nemotron-Parse-v1.1