Llama 3.2 1B is the smallest official Llama, designed for on-device deployment — it runs comfortably on a Snapdragon X Elite NPU, an Apple Neural Engine, or even a Raspberry Pi 5. Quality is naturally limited; it's best for summarization, classification, and tool-routing tasks rather than long-form generation. Expect 30-50 tok/s on midrange phones.
Llama 3.2 1B needs at minimum 0.6 GB of VRAM at Q4_K_M quantization (2.2 GB at FP16). The cheapest GPU that comfortably fits with KV-cache headroom is the Intel Arc A380 6GB (6 GB VRAM, $139 MSRP). Community benchmark submissions are open. This model fits a single consumer GPU under 48 GB, so a one-card build works.
Weights only. Add ~20-30% for KV cache at typical context lengths.
| Quant | VRAM | Quality | Notes |
|---|---|---|---|
| FP16 | 2.2 GB | Reference | Training-precision weights |
| Q8_0 | 1.1 GB | Near-lossless | 8-bit, ~0.1% perplexity hit |
| Q6_K | 0.9 GB | Very high | 6-bit, near Q8 quality |
| Q5_K_M | 0.8 GB | High | Strong middle ground |
| Q4_K_M | 0.6 GB | Balanced (recommended) | Default for local deployments |
| Q4_0 | 0.6 GB | Legacy 4-bit | Older GGUF, kept for compatibility |
| Q3_K_M | 0.5 GB | Lossy | When VRAM is very tight |
| Q2_K | 0.4 GB | Extreme | Rescue option, quality degrades visibly |
Filtered from MyAIHardware's GPU database. FP16 needs 2.2 GB; Q4_K_M needs 0.6 GB.
We don't have community benchmarks for this exact model yet.
Official weights + popular runtime tags.
ollama pull llama3.2:1blms get bartowski/Llama-3.2-1B-Instruct-GGUFStep-by-step guides that use this model.
Models in a similar size or capability class.
Use these in papers, blog posts, or internal docs.
@misc{llama-3-2-1b-2024,
title={Llama 3.2 1B},
author={Meta AI},
year={2024},
url={https://huggingface.co/meta-llama/Llama-3.2-1B-Instruct}
}Meta AI (2024). Llama 3.2 1B [Model card]. Hugging Face. https://huggingface.co/meta-llama/Llama-3.2-1B-Instruct
Llama 3.2 1B (Llama 3.2, Meta AI, 2024). Available at https://huggingface.co/meta-llama/Llama-3.2-1B-Instruct.