# Can I run Llama 4 Scout on Apple M3 Pro (18GB)?

> **No, not enough memory.** Needs ~64.2 GB even at Q4_K_M, but only ~12 GB is usable.

Last validated: 2026-06-15.

## Memory math
- Llama 4 Scout: 109B (MoE, 17B active); Q4_K_M weighs 60.87 GB on disk.
- At Q4_K_M with a 4k context it needs ~64.2 GB (weights + KV cache + ~0.8 GB runtime overhead, ±15% with context length).
- Apple M3 Pro (18GB): 18 GB unified, ~12 GB usable for model weights.
- Headroom: ~-52.2 GB.

## How to run
Recommended quant: n/a. Best tool on macOS: LM Studio.

```sh
ollama run llama4:scout
```

## Run Llama 4 Scout on other devices
- [Apple M4 Max (128GB)](https://localmodel.run/can-i-run/llama-4-scout/apple-m4-max-128gb)
- [Apple M5 Max (128GB)](https://localmodel.run/can-i-run/llama-4-scout/apple-m5-max-128gb)
- [Apple M3 Ultra (256GB)](https://localmodel.run/can-i-run/llama-4-scout/apple-m3-ultra-256gb)

## Other models that run on Apple M3 Pro (18GB)
- [Phi-4 14B (14B)](https://localmodel.run/can-i-run/phi-4-14b/apple-m3-18gb)
- [Qwen2.5 14B (14B)](https://localmodel.run/can-i-run/qwen2.5-14b/apple-m3-18gb)
- [Qwen3 14B (14B)](https://localmodel.run/can-i-run/qwen3-14b/apple-m3-18gb)
- [DeepSeek-R1-Distill-Qwen 14B (14B)](https://localmodel.run/can-i-run/deepseek-r1-distill-qwen-14b/apple-m3-18gb)
- [Qwen2.5 Coder 14B (14B)](https://localmodel.run/can-i-run/qwen2.5-coder-14b/apple-m3-18gb)
- [Mistral Nemo 12B (12.2B)](https://localmodel.run/can-i-run/mistral-nemo-12b/apple-m3-18gb)

Estimate, not a guarantee. Sources: https://ollama.com/library/llama4, https://ollama.com/library/llama4/tags, https://huggingface.co/unsloth/Llama-4-Scout-17B-16E-Instruct-GGUF, https://huggingface.co/meta-llama/Llama-4-Scout-17B-16E-Instruct.
More: https://localmodel.run/can-i-run/llama-4-scout/apple-m3-18gb