# Nemotron 3.5 Lightning 30B-A3B

> Canonical: https://www.overmindlab.ai/models/nvidia-nemotron-3-5-lightning-30b-a3b

NVIDIA's 30B/3B-active hybrid MoE for high-volume agent execution.

| Property | Value |
| --- | --- |
| Model ID | nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B |
| Provider | NVIDIA |
| Group | Nemotron 3.5 |
| Parameters | 30B-A3B |
| Tier | Large |
| Context window | 256K |
| Max training context | 256K |
| Tool calling | Yes |
| Training methods | — |
| Training cost | from $2.00 |
| Serving cost (per 1M output tokens) | from $6.00 |

## Overview

LoRA-only on Overmind via Unsloth. Native 256K context, Qwen-style tools, thinking off by default. Full FT disabled.

Nemotron 3.5 Lightning is a 30B total / 3B active hybrid reasoning MoE for tool calls, validation, and subagent work in long-running agents.

## Good for

- High-frequency tool-calling LoRA
- Output validation and formatting specialists
- Subagent routing in long-running agents

Train this model on your agent's data: https://docs.overmindlab.ai/models/training.md