Can it run?
Runs well
Using Q4_K_M in Ollama
Fit status
Runs well
Decode
52.9 tok/s
TTFT
3658 ms
Safe context
37K
Memory
11.2 GB / 25.9 GB
| Workload | Grade | Fit | Decode | TTFT | Context |
|---|---|---|---|---|---|
| Agentic Coding | C | Runs well | 52.9 tok/s | 5320 ms | 67K |
| Chat | C | Runs well | 52.9 tok/s | 1995 ms | 19K |
| Coding | C | Runs well | 52.9 tok/s | 3658 ms | 37K |
| RAG | C | Runs well | 52.9 tok/s | 6650 ms | 67K |
| Reasoning | C | Runs well | 52.9 tok/s | 4323 ms | 37K |
How Meta Llama 3.1 8B Instruct (8B params) fits at each quantization level on MacBook Pro M4 Max 36GB (25.9 GB usable).
| Quant | Bits | VRAM | Quality | Fit |
|---|---|---|---|---|
Q2_K | 2 | 3.1 GB | Low | D32 |
Q3_K_S | 3 | 3.9 GB | Low | D33 |
NVFP4 | 4 | 4.5 GB | Medium | D33 |
Q4_K_M | 4 | 4.9 GB | Medium | D34 |
Q5_K_M | 5 | 5.8 GB | High | D34 |
Q6_K | 6 | 6.6 GB | High | D35 |
Q8_0 | 8 | 8.6 GB | Very High | D37 |
F16Best for your GPU | 16 | 16.4 GB | Maximum | C43 |
