RTX 5090 32GBBudget pickB166.9 tok/s decode
~$1,999 MSRP
Can it run?
Runs with offload
Using Q4_K_M in Ollama
Fit status
Runs with offload
Decode
47.5 tok/s
TTFT
4080 ms
Safe context
17K
Memory
23.0 GB / 24.0 GB
| Workload | Grade | Fit | Decode | TTFT | Context |
|---|---|---|---|---|---|
| Agentic Coding | C | Runs with offload | 47.5 tok/s | 5934 ms | 33K |
| Chat | C | Runs with offload | 47.5 tok/s | 2225 ms | 8K |
| Coding | C | Runs with offload | 47.5 tok/s | 4080 ms | 17K |
| RAG | C | Runs with offload | 47.5 tok/s | 7418 ms | 33K |
| Reasoning | C | Runs with offload | 47.5 tok/s | 4822 ms | 17K |
How Qwen3-Coder 30B A3B Instruct (30.5B params) fits at each quantization level on RTX 4500 Ada 24GB (24.0 GB usable).
| Quant | Bits | VRAM | Quality | Fit |
|---|---|---|---|---|
Q2_K | 2 | 11.9 GB | Low | C40 |
Q3_K_S | 3 | 14.9 GB | Low | C43 |
NVFP4Best for your GPU | 4 | 17.1 GB | Medium | C45 |
Q4_K_M | 4 | 18.6 GB | Medium | C45 |
Q5_K_M | 5 | 22.0 GB | High | C45 |
Q6_K | 6 | 25.0 GB | High | F0 |
Q8_0 | 8 | 32.6 GB | Very High | F0 |
F16 | 16 | 62.5 GB | Maximum | F0 |
ollama run qwen-3-coder-30b-a3bhuggingface-cli download qwen-3-coder-30b-a3bUpgrade options
RTX 5090 32GBBudget pick~$1,999 MSRP
~$2,499 MSRP
RTX PRO 4500 Blackwell 32GBNVIDIA upgrade