Can it run?
Runs well
Using Q4_K_M in Ollama
Fit status
Runs well
Decode
101.0 tok/s
TTFT
1917 ms
Safe context
22K
Memory
4.4 GB / 6.0 GB
| Workload | Grade | Fit | Decode | TTFT | Context |
|---|---|---|---|---|---|
| Agentic Coding | B | Runs well | 101.0 tok/s | 2789 ms | 42K |
| Chat | B | Runs well | 101.0 tok/s | 1046 ms | 11K |
| Coding | B | Runs well | 101.0 tok/s | 1917 ms | 22K |
| RAG | B | Runs well | 101.0 tok/s | 3486 ms | 42K |
| Reasoning | B | Runs well | 101.0 tok/s | 2266 ms | 22K |
How Llama 3.2 3B (3B params) fits at each quantization level on GTX 1660 Super 6GB (6.0 GB usable).
| Quant | Bits | VRAM | Quality | Fit |
|---|---|---|---|---|
Q2_K | 2 | 1.2 GB | Low | D32 |
Q3_K_S | 3 | 1.5 GB | Low | D34 |
NVFP4 | 4 | 1.7 GB | Medium | D34 |
Q4_K_M | 4 | 1.8 GB | Medium | D35 |
Q5_K_M | 5 | 2.2 GB | High | D36 |
Q6_K | 6 | 2.5 GB | High | D37 |
Q8_0Best for your GPU | 8 | 3.2 GB | Very High | D40 |
F16 | 16 | 6.1 GB | Maximum | F0 |
ollama run llama-3.2-3bhuggingface-cli download llama-3.2-3b