Can it run?
Runs well
Using Q6_K in Ollama
Fit status
Runs well
Decode
200.8 tok/s
TTFT
964 ms
Safe context
29K
Memory
4.4 GB / 8.0 GB
| Workload | Grade | Fit | Decode | TTFT | Context |
|---|---|---|---|---|---|
| Agentic Coding | C | Runs well | 200.8 tok/s | 1403 ms | 58K |
| Chat | C | Runs well | 200.8 tok/s | 526 ms | 14K |
| Coding | C | Runs well | 200.8 tok/s | 964 ms | 29K |
| RAG | C | Runs well | 200.8 tok/s | 1753 ms | 58K |
| Reasoning | C | Runs well | 200.8 tok/s | 1140 ms | 29K |
How gemma 2 2b it (2B params) fits at each quantization level on RTX 3070 8GB (8.0 GB usable).
| Quant | Bits | VRAM | Quality | Fit |
|---|---|---|---|---|
Q2_K | 2 | 0.8 GB | Low | D32 |
Q3_K_S | 3 | 1.0 GB | Low | D32 |
NVFP4 | 4 | 1.1 GB | Medium | D32 |
Q4_K_M | 4 | 1.2 GB | Medium | D33 |
Q5_K_M | 5 | 1.4 GB | High | D33 |
Q6_K | 6 | 1.6 GB | High | D34 |
Q8_0 | 8 | 2.1 GB | Very High | D35 |
F16Best for your GPU | 16 | 4.1 GB | Maximum | C41 |
huggingface-cli download hf-bartowski--gemma-2-2b-it-gguf