GPUs & Macs › Apple M4 Max
Apple M4 Max: local LLM speed
Llama 2 7B (Q4_0) generates 75.0 tokens/s on the Apple M4 Max (Vulkan), and processes prompts at 725 tokens/s.
Measured results
| Model | Generate tok/s | Prompt tok/s | Quant | Backend | Test | Source |
|---|---|---|---|---|---|---|
| Llama 2 7B | 75.0 | 725 | Q4_0 | Vulkan FA off | tg128 | llama.cpp #10879 details |
| Llama 2 7B | 71.8 | 557 | Q4_0 | Vulkan FA on | tg128 | llama.cpp #10879 details |
| gpt-oss 20B | 92.4 | 1,277 | MXFP4 | Metal FA on | tg128 | llama.cpp #15396 details |
What fits on the Apple M4 Max
Estimates for this chip are scaled from a measured run: Llama 2 7B Q4_0 at 75.0 tok/s (Vulkan, source). Fit assumes the whole model in GPU memory with ~8K context.
With 16 GB unified memory
| Model | Best quant that fits | File | Expected tok/s |
|---|---|---|---|
| Qwen3.8-27B | UD-IQ2_XXS | 7.3 GB | 39.5 estimate |
| gemma-4-12B-it | Q6_K | 9.8 GB | 29.3 estimate |
| Qwen3.5-9B | Q8_0 | 9.5 GB | 30.1 estimate |
| Qwen3.5-4B | Q8_0 | 4.5 GB | 64.0 estimate |
| Qwen3.6-35B-A3B | UD-IQ1_M | 10.0 GB | 167 estimate |
| gemma-4-26B-A4B-it | UD-IQ2_M | 10.0 GB | 93.1 estimate |
| MiMo-V2.6-Distill-Qwen-9B | Q8_0 | 9.5 GB | 30.0 estimate |
| gemma-4-E4B-it | Q8_0 | 8.2 GB | 35.0 estimate |
With 36 GB unified memory
| Model | Best quant that fits | File | Expected tok/s |
|---|---|---|---|
| Qwen3.8-27B | UD-Q6_K | 22.0 GB | 13.0 estimate |
| gemma-4-12B-it | Q8_0 | 12.7 GB | 22.6 estimate |
| Qwen3.5-9B | Q8_0 | 9.5 GB | 30.1 estimate |
| Qwen3.5-4B | Q8_0 | 4.5 GB | 64.0 estimate |
| Qwen3.6-35B-A3B | UD-Q5_K_S | 24.9 GB | 67.1 estimate |
| gemma-4-26B-A4B-it | UD-Q6_K | 23.2 GB | 40.2 estimate |
| MiMo-V2.6-Distill-Qwen-9B | Q8_0 | 9.5 GB | 30.0 estimate |
| gemma-4-E4B-it | Q8_0 | 8.2 GB | 35.0 estimate |
With 64 GB unified memory
| Model | Best quant that fits | File | Expected tok/s |
|---|---|---|---|
| Qwen3.8-27B | Q8_0 | 29.0 GB | 9.9 estimate |
| gemma-4-12B-it | Q8_0 | 12.7 GB | 22.6 estimate |
| Qwen3.5-9B | Q8_0 | 9.5 GB | 30.1 estimate |
| Qwen3.5-4B | Q8_0 | 4.5 GB | 64.0 estimate |
| Qwen3.6-35B-A3B | Q8_0 | 36.9 GB | 45.3 estimate |
| gemma-4-26B-A4B-it | Q8_0 | 26.9 GB | 34.7 estimate |
| MiMo-V2.6-Distill-Qwen-9B | Q8_0 | 9.5 GB | 30.0 estimate |
| gemma-4-E4B-it | Q8_0 | 8.2 GB | 35.0 estimate |
With 128 GB unified memory
| Model | Best quant that fits | File | Expected tok/s |
|---|---|---|---|
| Qwen3.8-27B | Q8_0 | 29.0 GB | 9.9 estimate |
| gemma-4-12B-it | Q8_0 | 12.7 GB | 22.6 estimate |
| Qwen3.5-9B | Q8_0 | 9.5 GB | 30.1 estimate |
| Qwen3.5-4B | Q8_0 | 4.5 GB | 64.0 estimate |
| Qwen3.6-35B-A3B | Q8_0 | 36.9 GB | 45.3 estimate |
| Qwen3.8-Flash-Next | UD-IQ3_XXS | 82.0 GB | 3.5 estimate |
| gemma-4-26B-A4B-it | Q8_0 | 26.9 GB | 34.7 estimate |
| MiMo-V2.6-Distill-Qwen-9B | Q8_0 | 9.5 GB | 30.0 estimate |
Similar hardware
Apple M1 (7-core GPU) · Apple M1 (8-core GPU) · Apple M1 Pro (14-core GPU) · Apple M1 Pro (16-core GPU) · Apple M1 Max (24-core GPU) · Apple M1 Max (32-core GPU) · Apple M1 Ultra (48-core GPU) · Apple M1 Ultra (64-core GPU)