GPUs & Macs › Apple M1 Max
Apple M1 Max: local LLM speed
2 measured results for the Apple M1 Max.
Measured results
| Model | Generate tok/s | Prompt tok/s | Quant | Backend | Test | Source |
|---|---|---|---|---|---|---|
| gpt-oss 20B | 75.2 | 995 | MXFP4 | Metal FA on | tg128 | llama.cpp #15396 details |
| gpt-oss 20B | 75.2 | 995 | MXFP4 | Metal,BLAS FA on | tg128 | llama.cpp #15396 details |
What fits on the Apple M1 Max
No measured run yet; estimates use spec bandwidth × 0.83 (typical efficiency for Apple). 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 | 45.7 estimate |
| gemma-4-12B-it | Q6_K | 9.8 GB | 34.0 estimate |
| Qwen3.5-9B | Q8_0 | 9.5 GB | 34.9 estimate |
| Qwen3.5-4B | Q8_0 | 4.5 GB | 74.2 estimate |
| Qwen3.6-35B-A3B | UD-IQ1_M | 10.0 GB | 193 estimate |
| gemma-4-26B-A4B-it | UD-IQ2_M | 10.0 GB | 108 estimate |
| MiMo-V2.6-Distill-Qwen-9B | Q8_0 | 9.5 GB | 34.8 estimate |
| gemma-4-E4B-it | Q8_0 | 8.2 GB | 40.6 estimate |
With 36 GB unified memory
| Model | Best quant that fits | File | Expected tok/s |
|---|---|---|---|
| Qwen3.8-27B | UD-Q6_K | 22.0 GB | 15.1 estimate |
| gemma-4-12B-it | Q8_0 | 12.7 GB | 26.2 estimate |
| Qwen3.5-9B | Q8_0 | 9.5 GB | 34.9 estimate |
| Qwen3.5-4B | Q8_0 | 4.5 GB | 74.2 estimate |
| Qwen3.6-35B-A3B | UD-Q5_K_S | 24.9 GB | 77.8 estimate |
| gemma-4-26B-A4B-it | UD-Q6_K | 23.2 GB | 46.6 estimate |
| MiMo-V2.6-Distill-Qwen-9B | Q8_0 | 9.5 GB | 34.8 estimate |
| gemma-4-E4B-it | Q8_0 | 8.2 GB | 40.6 estimate |
With 64 GB unified memory
| Model | Best quant that fits | File | Expected tok/s |
|---|---|---|---|
| Qwen3.8-27B | Q8_0 | 29.0 GB | 11.4 estimate |
| gemma-4-12B-it | Q8_0 | 12.7 GB | 26.2 estimate |
| Qwen3.5-9B | Q8_0 | 9.5 GB | 34.9 estimate |
| Qwen3.5-4B | Q8_0 | 4.5 GB | 74.2 estimate |
| Qwen3.6-35B-A3B | Q8_0 | 36.9 GB | 52.6 estimate |
| gemma-4-26B-A4B-it | Q8_0 | 26.9 GB | 40.2 estimate |
| MiMo-V2.6-Distill-Qwen-9B | Q8_0 | 9.5 GB | 34.8 estimate |
| gemma-4-E4B-it | Q8_0 | 8.2 GB | 40.6 estimate |
With 128 GB unified memory
| Model | Best quant that fits | File | Expected tok/s |
|---|---|---|---|
| Qwen3.8-27B | Q8_0 | 29.0 GB | 11.4 estimate |
| gemma-4-12B-it | Q8_0 | 12.7 GB | 26.2 estimate |
| Qwen3.5-9B | Q8_0 | 9.5 GB | 34.9 estimate |
| Qwen3.5-4B | Q8_0 | 4.5 GB | 74.2 estimate |
| Qwen3.6-35B-A3B | Q8_0 | 36.9 GB | 52.6 estimate |
| Qwen3.8-Flash-Next | UD-IQ3_XXS | 82.0 GB | 4.1 estimate |
| gemma-4-26B-A4B-it | Q8_0 | 26.9 GB | 40.2 estimate |
| MiMo-V2.6-Distill-Qwen-9B | Q8_0 | 9.5 GB | 34.8 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)