Can it run?
Runs well
Using Q4_K_M in Ollama
Fit status
Runs well
Decode
68.3 tok/s
TTFT
2833 ms
Safe context
20K
Memory
4.9 GB / 6.0 GB
| Workload | Grade | Fit | Decode | TTFT | Context |
|---|---|---|---|---|---|
| Agentic Coding | C | Tight fit | 68.3 tok/s | 4121 ms | 36K |
| Chat | B | Runs well | 68.3 tok/s | 1545 ms | 10K |
| Coding | B | Runs well | 68.3 tok/s | 2833 ms | 20K |
| RAG | C | Tight fit | 68.3 tok/s | 5151 ms | 36K |
| Reasoning | B | Runs well | 68.3 tok/s | 3348 ms | 20K |
How Phi 3 Mini 3.8B (3.799999952316284B params) fits at each quantization level on GTX 1660 Ti 6GB (6.0 GB usable).
| Quant | Bits | VRAM | Quality | Fit |
|---|---|---|---|---|
Q2_K | 2 | 1.5 GB | Low | D34 |
Q3_K_S | 3 | 1.9 GB | Low | D35 |
NVFP4 | 4 | 2.1 GB | Medium | D36 |
Q4_K_M | 4 | 2.3 GB | Medium | D37 |
Q5_K_M | 5 | 2.7 GB | High | D38 |
Q6_KBest for your GPU | 6 | 3.1 GB | High | D39 |
Q8_0 | 8 | 4.1 GB | Very High | C43 |
F16 | 16 | 7.8 GB | Maximum | F0 |
ollama run phi-3-mini-3.8bhuggingface-cli download phi-3-mini-3.8b