# Llama-3.2-1B-Instruct GPU requirements

Llama-3.2-1B-Instruct has 1.2B parameters. To serve it, plan for about 3.0 GB of GPU memory in 16-bit, 1.5 GB in 8-bit or 0.7 GB in 4-bit, including about 20% for the KV cache and activations.

[Deploy it on Arrakis](/console/#/models?model=meta-llama/Llama-3.2-1B-Instruct).

## What it takes to serve it

| Precision | Memory and the smallest setup |
|---|---|
| 16-bit | 3.0 GB: one [RTX 4090](/gpus/rtx-4090.html.md) or [RTX 3090](/gpus/rtx-3090.html.md), RTX 3090 from $0.19 per GPU-hour |
| 8-bit | 1.5 GB: one [RTX 4090](/gpus/rtx-4090.html.md) or [RTX 3090](/gpus/rtx-3090.html.md), RTX 3090 from $0.19 per GPU-hour |
| 4-bit | 0.7 GB: one [RTX 4090](/gpus/rtx-4090.html.md) or [RTX 3090](/gpus/rtx-3090.html.md), RTX 3090 from $0.19 per GPU-hour |

## What it takes to fine-tune it

| Method | Memory and the smallest setup |
|---|---|
| QLoRA | 0.9 GB: one [RTX 4090](/gpus/rtx-4090.html.md) or [RTX 3090](/gpus/rtx-3090.html.md), RTX 3090 from $0.19 per GPU-hour |
| LoRA | 3.0 GB: one [RTX 4090](/gpus/rtx-4090.html.md) or [RTX 3090](/gpus/rtx-3090.html.md), RTX 3090 from $0.19 per GPU-hour |
| Full fine-tuning | 20 GB: one [RTX 4090](/gpus/rtx-4090.html.md) or [RTX 3090](/gpus/rtx-3090.html.md), RTX 3090 from $0.19 per GPU-hour |

## About the model

| | |
|---|---|
| Parameters | `1.2B` |
| Published in | `BF16` |
| License | `llama3.2` |
| Released | `2024-09-18` |
| Weights | [meta-llama/Llama-3.2-1B-Instruct](https://huggingface.co/meta-llama/Llama-3.2-1B-Instruct) |

## Serve it yourself

```
vllm serve meta-llama/Llama-3.2-1B-Instruct --tensor-parallel-size 1
```

## Questions

**How much GPU memory does Llama-3.2-1B-Instruct need?** To serve it, about 3 GB in 16-bit, 1 GB in 8-bit, 1 GB in 4-bit, including about 20% for the KV cache and activations. To fine-tune it, about 1 GB with QLoRA, 3 GB with LoRA and 20 GB for full fine-tuning with Adam, before activations.

**Can I run Llama-3.2-1B-Instruct on one GPU?** Yes. In 16-bit it needs about 3.0 GB, which fits on one RTX 4090 or RTX 3090.

**What does it cost to run Llama-3.2-1B-Instruct?** In 16-bit it needs one RTX 4090 or RTX 3090; in 4-bit, one RTX 4090 or RTX 3090. Each setup on this page shows the lowest price per GPU-hour on Arrakis right now, and you pay by the hour from a prepaid balance.

**How do I deploy Llama-3.2-1B-Instruct on Arrakis?** Open Models in the Arrakis console, paste meta-llama/Llama-3.2-1B-Instruct, choose vLLM, SGLang or llama.cpp, and rent a GPU it fits on. You get an OpenAI-compatible endpoint with its own API key, and root SSH to the machine.


More models: [/models](/models.html.md).