Can the Apple M2 Pro (32GB) run Mixtral 8x7B?
- FP16
- Q8_0
- Q6_K
- Q5_K_M
- Q4_K_M
- Q3_K_M
- Q2_K
Only just. Mixtral 8x7B fits on the Apple M2 Pro (32GB) at Q2_K, which costs noticeable quality. Expect about 25 tokens/second. A card with more memory is a better experience.
Run it with Ollama
Ollama's default mixtral:8x7b download is Q4_K_M, which needs about 30 GB, more than the 21.4 GB available. Download the Q2_K GGUF instead (LM Studio lists quants per model), or try a smaller context length.
Run it in the cloud
Want full quality? Run Mixtral 8x7B on a bigger rented GPU.
Rent a cloud GPU by the hour. No upfront cost, stop any time.
Memory needed by quantization
At 8,192 tokens of context. The black line marks the 21.4 GB the Apple M2 Pro (32GB) can use. macOS lets the GPU use about 21.4 GB of the 32 GB unified memory by default.
| Quant | Quality | Memory | Fits? | Speed | |
|---|---|---|---|---|---|
| FP16 | Full quality | 95.1 GB | Too big | — | |
| Q8_0 | Near-lossless | 51.3 GB | Too big | — | |
| Q6_K | Excellent | 40 GB | Too big | — | |
| Q5_K_M | Very good | 34.9 GB | Too big | — | |
| Q4_K_M | Good, most popular | 30 GB | Too big | — | |
| Q3_K_M | Noticeably degraded | 24.5 GB | Too big | — | |
| Q2_K | Heavily degraded | 19.2 GB | Fits | ~25 tok/s |
GPUs that run Mixtral 8x7B comfortably
Frequently asked questions
Can Apple M2 Pro (32GB) run Mixtral 8x7B?
Only just. Mixtral 8x7B fits on the Apple M2 Pro (32GB) at Q2_K, which costs noticeable quality. Expect about 25 tokens/second. A card with more memory is a better experience.
How much VRAM does Mixtral 8x7B need?
About 30 GB at Q4_K_M with 8,192 tokens of context, and about 95.1 GB at full FP16 precision.
How fast is Mixtral 8x7B on the Apple M2 Pro (32GB)?
Roughly 25 tokens per second at Q2_K, estimated from the Apple M2 Pro (32GB)'s 200 GB/s memory bandwidth.