# Llama 3.2 3B Instruct

> Canonical: https://www.overmindlab.ai/models/llama-3-2-3b-instruct

Compact Llama 3.2 step up, still trainable at the full 128K context.

| Property | Value |
| --- | --- |
| Model ID | meta-llama/Llama-3.2-3B-Instruct |
| Provider | Meta |
| Group | Llama 3 |
| Parameters | 3B |
| Tier | Compact |
| Context window | 128K |
| Max training context | 128K |
| Tool calling | No |
| Training methods | — |
| Training cost | from $0.55 |
| Serving cost (per 1M output tokens) | from $4.00 |

## Overview

Llama 3 on Overmind covers Meta's instruct models at 1B, 3B, and 8B, each trainable at their full 128K context. Pick this family when your rows are long and you still want a compact model you can serve cheaply after training.

Meta's instruct models, all three trainable at their full 128K context. The natural choice when your rows are long and you want a compact model.

## Good for

- Long-row training on compact bases
- Meta-stack compatibility
- Tool-calling agents at 1B through 8B

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

## Related models

- [Llama 3.2 1B Instruct](https://www.overmindlab.ai/models/llama-3-2-1b-instruct): 1B, 128K context
- [Llama 3.1 8B Instruct](https://www.overmindlab.ai/models/meta-llama-3-1-8b-instruct-reference): 8B, 128K context
- [Llama 3.1 8B Instruct](https://www.overmindlab.ai/models/llama-3-1-8b-instruct): 8B, 128K context
- [Llama 3.3 70B Instruct](https://www.overmindlab.ai/models/llama-3-3-70b-instruct): 70B, 128K context