Can it run?
Runs well
Using Q4_K_M in Ollama
Fit status
Runs well
Decode
64.0 tok/s
TTFT
3025 ms
Safe context
21K
Memory
3.0 GB / 4.0 GB
| Workload | Grade | Fit | Decode | TTFT | Context |
|---|---|---|---|---|---|
| Agentic Coding | C | Runs well | 64.0 tok/s | 4400 ms | 43K |
| Chat | C | Runs well | 64.0 tok/s | 1650 ms | 11K |
| Coding | C | Runs well | 64.0 tok/s | 3025 ms | 21K |
| RAG | C | Runs well | 64.0 tok/s | 5500 ms | 43K |
| Reasoning | C | Runs well | 64.0 tok/s | 3575 ms | 21K |
How Llama 3.2 1B (1B params) fits at each quantization level on GTX 1650 4GB (4.0 GB usable).
| Quant | Bits | VRAM | Quality | Fit |
|---|---|---|---|---|
Q2_K | 2 | 0.4 GB | Low | D30 |
Q3_K_S | 3 | 0.5 GB | Low | D31 |
NVFP4 | 4 | 0.6 GB | Medium | D31 |
Q4_K_M | 4 | 0.6 GB | Medium | D31 |
Q5_K_M | 5 | 0.7 GB | High | D32 |
Q6_K | 6 | 0.8 GB | High | D32 |
Q8_0Best for your GPU | 8 | 1.1 GB | Very High | D34 |
F16 | 16 | 2.1 GB | Maximum | D39 |
ollama run llama-3.2-1bhuggingface-cli download llama-3.2-1b