MacBook Pro M4 32GBBest value
C5.9 tok/s decode
~$799 MSRP
Can it run?
Runs with offload
Using Q4_K_M in Ollama
Fit status
Runs with offload
Decode
36.9 tok/s
TTFT
5249 ms
Safe context
15K
Memory
21.6 GB / 20.0 GB
Offload
10%
| Workload | Grade | Fit | Decode | TTFT | Context |
|---|---|---|---|---|---|
| Agentic Coding | F | Too heavy | 32.8 tok/s | 8589 ms | 25K |
| Chat | C | Runs with offload | 32.8 tok/s | 3221 ms | 8K |
| Coding | C | Runs with offload | 36.9 tok/s | 5249 ms | 15K |
| RAG | F | Too heavy | 32.8 tok/s | 10736 ms | 25K |
| Reasoning | C | Runs with offload (needs ~1.1 GB host RAM) | 30.7 tok/s | 7444 ms | 15K |
How Devstral Small 2 24B Instruct (24B params) fits at each quantization level on RX 7900 XT 20GB (20.0 GB usable).
| Quant | Bits | VRAM | Quality | Fit |
|---|---|---|---|---|
Q2_K | 2 | 9.4 GB | Low | D40 |
Q3_K_S | 3 | 11.8 GB | Low | C42 |
NVFP4 | 4 |
ollama run devstral-small-2-24bhuggingface-cli download devstral-small-2-24bUpgrade options
~$799 MSRP
~$999 MSRP
RTX 5090 32GBBiggest leap13.4 GB |
| Medium |
| C44 |
Q4_K_MBest for your GPU | 4 | 14.6 GB | Medium | C44 |
Q5_K_M | 5 | 17.3 GB | High | C44 |
Q6_K | 6 | 19.7 GB | High | C44 |
Q8_0 | 8 | 25.7 GB | Very High | F0 |
F16 | 16 | 49.2 GB | Maximum | F0 |
~$1,999 MSRP