Can it run?
Runs well
Using Q4_K_M in Ollama
Fit status
Runs well
Decode
66.8 tok/s
TTFT
2897 ms
Safe context
16K
Memory
12.8 GB / 16.0 GB
| Workload | Grade | Fit | Decode | TTFT | Context |
|---|---|---|---|---|---|
| Agentic Coding | C | Tight fit | 77.0 tok/s | 3656 ms | 16K |
| Chat | B | Runs well | 77.0 tok/s | 1371 ms | 11K |
| Coding | B | Runs well | 66.8 tok/s | 2897 ms | 16K |
| RAG | C | Tight fit | 77.0 tok/s | 4569 ms | 16K |
| Reasoning | B | Runs well | 77.0 tok/s | 2970 ms | 16K |
How CodeLlama 13B Instruct (13B params) fits at each quantization level on RTX 4080 Super 16GB (16.0 GB usable).
| Quant | Bits | VRAM | Quality | Fit |
|---|---|---|---|---|
Q2_K | 2 | 5.1 GB | Low | D36 |
Q3_K_S | 3 | 6.4 GB | Low | D37 |
NVFP4 | 4 |
ollama run codellama-13b-instructhuggingface-cli download codellama-13b-instruct7.3 GB |
| Medium |
| D39 |
Q4_K_M | 4 | 7.9 GB | Medium | D39 |
Q5_K_M | 5 | 9.4 GB | High | C41 |
Q6_KBest for your GPU | 6 | 10.7 GB | High | C43 |
Q8_0 | 8 | 13.9 GB | Very High | C44 |
F16 | 16 | 26.7 GB | Maximum | F0 |