# Can I run DeepSeek V3 on Apple M4 (16GB)?

> **No, not enough memory.** Needs ~383.7 GB even at Q4_K_M, but only ~10.5 GB is usable.

Last validated: 2026-06-15.

## Memory math
- DeepSeek V3: 671B (MoE, 37B active); Q4_K_M weighs 376.66 GB on disk.
- At Q4_K_M with a 4k context it needs ~383.7 GB (weights + KV cache + ~0.8 GB runtime overhead, ±15% with context length).
- Apple M4 (16GB): 16 GB unified, ~10.5 GB usable for model weights.
- Headroom: ~-373.2 GB.

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

```sh
ollama run deepseek-v3:671b
```

## Other models that run on Apple M4 (16GB)
- [Qwen2.5 Coder 14B (14B)](https://localmodel.run/can-i-run/qwen2.5-coder-14b/apple-m4-16gb)
- [Mistral Nemo 12B (12.2B)](https://localmodel.run/can-i-run/mistral-nemo-12b/apple-m4-16gb)
- [Gemma 3 12B (12B)](https://localmodel.run/can-i-run/gemma-3-12b/apple-m4-16gb)
- [Llama 3.2 Vision 11B (10.7B)](https://localmodel.run/can-i-run/llama-3.2-vision-11b/apple-m4-16gb)
- [Falcon3 10B (10B)](https://localmodel.run/can-i-run/falcon3-10b/apple-m4-16gb)
- [Gemma 2 9B (9B)](https://localmodel.run/can-i-run/gemma-2-9b/apple-m4-16gb)

Estimate, not a guarantee. Sources: https://ollama.com/library/deepseek-v3, https://ollama.com/library/deepseek-v3/tags, https://huggingface.co/unsloth/DeepSeek-V3-0324-GGUF, https://huggingface.co/deepseek-ai/DeepSeek-V3-0324.
More: https://localmodel.run/can-i-run/deepseek-v3/apple-m4-16gb