Can it run?
Runs well
Using Q4_K_M in Ollama
Fit status
Runs well
Decode
109.5 tok/s
TTFT
1768 ms
Safe context
63K
Memory
6.6 GB / 25.9 GB
| Workload | Grade | Fit | Decode | TTFT | Context |
|---|---|---|---|---|---|
| Agentic Coding | C | Runs well | 109.5 tok/s | 2572 ms | 126K |
| Chat | C | Runs well | 109.5 tok/s | 964 ms | 32K |
| Coding | C | Runs well | 109.5 tok/s | 1768 ms | 63K |
| RAG | C | Runs well | 109.5 tok/s | 3215 ms | 126K |
| Reasoning | C | Runs well | 109.5 tok/s | 2089 ms | 63K |
How TinyLlama 1.1B Chat v1.0 (1.100000023841858B params) fits at each quantization level on MacBook Pro M3 Pro 36GB (25.9 GB usable).
| Quant | Bits | VRAM | Quality | Fit |
|---|---|---|---|---|
Q2_K | 2 | 0.4 GB | Low | D30 |
Q3_K_S | 3 | 0.5 GB | Low | D30 |
NVFP4 | 4 |
0.6 GB |
| Medium |
| D30 |
Q4_K_M | 4 | 0.7 GB | Medium | D30 |
Q5_K_M | 5 | 0.8 GB | High | D30 |
Q6_K | 6 | 0.9 GB | High | D30 |
Q8_0 | 8 | 1.2 GB | Very High | D30 |
F16Best for your GPU | 16 | 2.3 GB | Maximum | D31 |
