RTX 5070 12GBBudget pickB89 tok/s decode
~$549 MSRP
Can it run?
Runs well
Using Q4_K_M in Ollama
Fit status
Runs well
Decode
84.2 tok/s
TTFT
2301 ms
Safe context
21K
Memory
8.3 GB / 11.0 GB
| Workload | Grade | Fit | Decode | TTFT | Context |
|---|---|---|---|---|---|
| Agentic Coding | C | Tight fit | 84.2 tok/s | 3346 ms | 37K |
| Chat | B | Runs well | 84.2 tok/s | 1255 ms | 11K |
| Coding | B | Runs well | 84.2 tok/s | 2301 ms | 21K |
| RAG | C | Tight fit | 84.2 tok/s | 4183 ms | 37K |
| Reasoning | B | Runs well | 84.2 tok/s | 2719 ms | 21K |
How EXAONE 3.5 7.8B Instruct (7.800000190734863B params) fits at each quantization level on RTX 2080 Ti 11GB (11.0 GB usable).
| Quant | Bits | VRAM | Quality | Fit |
|---|---|---|---|---|
Q2_K | 2 | 3.0 GB | Low | D35 |
Q3_K_S | 3 | 3.8 GB | Low | D36 |
NVFP4 | 4 |
Upgrade options
RTX 5070 12GBBudget pick~$549 MSRP
RTX 3080 12GBBest value~$799 MSRP
RTX 3080 Ti 12GBBiggest leap~$1,199 MSRP
4.4 GB |
| Medium |
| D38 |
Q4_K_M | 4 | 4.8 GB | Medium | D38 |
Q5_K_M | 5 | 5.6 GB | High | D40 |
Q6_KBest for your GPU | 6 | 6.4 GB | High | C41 |
Q8_0 | 8 | 8.3 GB | Very High | C44 |
F16 | 16 | 16.0 GB | Maximum | F0 |