Can it run?
Runs well
Using Q4_K_M in llama.cpp
Fit status
Runs well
Decode
86.8 tok/s
TTFT
2232 ms
Safe context
8K
Memory
8.2 GB / 12.0 GB
| Workload | Grade | Fit | Decode | TTFT | Context |
|---|---|---|---|---|---|
| Agentic Coding | B | Runs well | 86.8 tok/s | 3246 ms | 8K |
| Chat | B | Runs well | 86.8 tok/s | 1217 ms | 8K |
| Coding | B | Runs well | 86.8 tok/s | 2232 ms | 8K |
| RAG | B | Runs well | 86.8 tok/s | 4057 ms | 8K |
| Reasoning | B | Runs well | 86.8 tok/s | 2637 ms | 8K |
How Aya Expanse 8B (8B params) fits at each quantization level on RTX 5070 12GB (12.0 GB usable).
| Quant | Bits | VRAM | Quality | Fit |
|---|---|---|---|---|
Q2_K | 2 | 3.1 GB | Low | D35 |
Q3_K_S | 3 | 3.9 GB | Low | D36 |
NVFP4 | 4 | 4.5 GB | Medium | D37 |
Q4_K_M | 4 | 4.9 GB | Medium | D38 |
Q5_K_M | 5 | 5.8 GB | High | D40 |
Q6_K | 6 | 6.6 GB | High | C41 |
Q8_0Best for your GPU | 8 | 8.6 GB | Very High | C44 |
F16 | 16 | 16.4 GB | Maximum | F0 |
huggingface-cli download aya-expanse-8b