Skip to main content

Ctrl+K

📢 Notice: NIM LLM and VLM 3.0 is now available alongside NIM LLM and VLM 2.0. Consider 3.0 for workloads that depend on Dynamo-based distributed inference with independently scaled workers.

NVIDIA NIM for Large Language Models (LLM) and Vision Language Models (VLM)

Documentation Home

NVIDIA NIM for Large Language Models (LLM) and Vision Language Models (VLM)

Documentation Home

Table of Contents

About NIM LLM and VLM

Overview
NIM Offerings
Enterprise-Grade Inference Software Stack
Release Notes

Get Started

About Get Started
Prerequisites
Configuration
Installation
Quickstart
Advanced
- Get Started with DeepSeek-V4-Pro-0813

Deployment

Model Profiles and Selection
Model Download
Model-Free NIM
Kubernetes Deployment
Cloud Service Provider (CSP) Deployment
- Google Cloud
- AWS
- Azure
- Oracle
Air-Gap Deployment
Multi-Node Deployment
vGPU Deployment

Advanced Use Cases

Fine-Tuning with LoRA
Speculative Decoding
Payload Capture
Tool Calling and MCP Integration
Custom Parsers and Chat Templates
Custom Logits Processing
Prompt Embeddings
Image, Audio, and Video Input

AI Assistant Integrations

Use Claude Code with NIM
Use Codex CLI with NIM

Reference

Benchmarking
Architecture
Environment Variables
API Reference
CLI Reference
Advanced Configuration
Logging and Observability
Model Signature Verification
1.x Migration Guide
Support Matrix for NIMs
Archived Versions

Troubleshooting

GPU Memory (OOM) Errors
CUDA Driver Initialization
OpenShift SCC Admission Failures

Resources

Support and FAQ
Related Products
Legal

Is this page helpful?

Search

Ctrl+K

Last updated on Sep 11, 2026.