MacBook Pro M3 Max 128GBBest value
D3 tok/s decode
~$2,499 MSRP
Can it run?
Too heavy
Using Q4_K_M in vLLM
Fit status
Too heavy
Decode
10.5 tok/s
TTFT
18453 ms
Safe context
8K
Memory
101.4 GB / 48.0 GB
| Workload | Grade | Fit | Decode | TTFT | Context |
|---|---|---|---|---|---|
| Agentic Coding | F | Too heavy | 10.5 tok/s | 26840 ms | 13K |
| Chat | F | Too heavy | 10.5 tok/s | 10065 ms | 4K |
| Coding | F | Too heavy | 10.5 tok/s | 18453 ms | 8K |
| RAG | F | Too heavy | 10.5 tok/s | 33550 ms | 13K |
| Reasoning | F | Too heavy | 10.5 tok/s | 21808 ms | 8K |
How Devstral 2 123B Instruct (123B params) fits at each quantization level on RTX 6000 Ada 48GB (48.0 GB usable).
| Quant | Bits | VRAM | Quality | Fit |
|---|---|---|---|---|
Q2_K | 2 | 48.0 GB | Low | C45 |
Q3_K_S | 3 | 60.3 GB | Low | F0 |
NVFP4 | 4 | 68.9 GB | Medium | F0 |
Q4_K_M | 4 | 75.0 GB | Medium | F0 |
Q5_K_M | 5 | 88.6 GB | High | F0 |
Q6_K | 6 | 100.9 GB | High | F0 |
Q8_0 | 8 | 131.6 GB | Very High | F0 |
F16 | 16 | 252.2 GB | Maximum | F0 |
Upgrade options
~$2,499 MSRP
~$8,000 MSRP
NVIDIA B200 180GBBiggest leap~$30,000 MSRP
NVIDIA H200 141GBNVIDIA upgrade~$30,000 MSRP