Can the Apple M4 Max (128GB) run DeepSeek R1 Distill Qwen 32B?
Yes
QuantFP16Full quality
Memory68.3 GBof 96 GB
Speed~5tokens/s
Yes. The Apple M4 Max (128GB) can run DeepSeek R1 Distill Qwen 32B fully in unified memory at FP16 (full quality), using about 68.3 GB of the 96 GB available, at roughly 5 tokens/second.
Try other context lengths in the calculator →Memory needed by quantization
At 8,192 tokens of context. macOS lets the GPU use about 96 GB of the 128 GB unified memory by default.
| Quant | Quality | Memory | Fits? | Speed |
|---|---|---|---|---|
| FP16 | Full quality | 68.3 GB | Fits | ~5 tok/s |
| Q8_0 | Near-lossless | 37.6 GB | Fits | ~9 tok/s |
| Q6_K | Excellent | 29.6 GB | Fits | ~12 tok/s |
| Q5_K_M | Very good | 26.1 GB | Fits | ~14 tok/s |
| Q4_K_M | Good, most popular | 22.6 GB | Fits | ~16 tok/s |
| Q3_K_M | Noticeably degraded | 18.8 GB | Fits | ~20 tok/s |
| Q2_K | Heavily degraded | 15 GB | Fits | ~27 tok/s |
How to run DeepSeek R1 Distill Qwen 32B on the Apple M4 Max (128GB)
With Ollama installed, one command downloads and starts it:
ollama run deepseek-r1:32bOllama's default download is Q4_K_M (about 22.6 GB here). You have room for FP16, which is full quality: pick that GGUF in LM Studio or from Hugging Face.
GPUs that run DeepSeek R1 Distill Qwen 32B comfortably
- NVIDIA RTX 3090 (24 GB)
- AMD Radeon RX 7900 XTX (24 GB)
- NVIDIA RTX 4090 (24 GB)
- NVIDIA RTX 5090 (32 GB)
- Apple M4 Pro (48GB)
Models that run well on the Apple M4 Max (128GB)
Run it in the cloud
Don't want to buy hardware? Run DeepSeek R1 Distill Qwen 32B on a rented GPU today.
Rent a cloud GPU by the hour. No upfront cost, cancel any time.