C246 tok/s decode
~$1,999 MSRP
Can it run?
Runs well
Using Q4_K_M in Ollama
Fit status
Runs well
Decode
20.0 tok/s
TTFT
9690 ms
Safe context
37K
Memory
11.2 GB / 25.9 GB
| Workload | Grade | Fit | Decode | TTFT | Context |
|---|---|---|---|---|---|
| Agentic Coding | C | Runs well | 22.4 tok/s | 12550 ms | 67K |
| Chat | C | Runs well | 22.4 tok/s | 4706 ms | 19K |
| Coding | C | Runs well | 20.0 tok/s | 9690 ms | 37K |
| RAG | C | Runs well | 22.4 tok/s | 15687 ms | 67K |
| Reasoning | C | Runs well | 22.4 tok/s | 10197 ms | 37K |
How DeepSeek R1 0528 Qwen3 8B (8B params) fits at each quantization level on MacBook Pro M3 Pro 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 |
Upgrade options
~$1,999 MSRP
~$2,499 MSRP
4.5 GB |
| Medium |
| D33 |
Q4_K_M | 4 | 4.9 GB | Medium | D33 |
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 |
~$3,999 MSRP