Qwen3 30B-A3B (MoE): VRAM by quant
30.5B params · KV field 98304 · max context field 131072.
| Quant | Bits | Weight GB sketch |
|---|---|---|
| Q4_K_M | 4.85 bpw | ~18.9 GB |
| Q5_K_M | 5.7 bpw | ~22.0 GB |
| Q6_K | 6.6 bpw | ~25.3 GB |
| Q8_0 | 8.5 bpw | ~32.6 GB |
| MXFP4 | 4.25 bpw | ~16.8 GB |
| F16/BF16 | 16.0 bpw | ~61.0 GB |
Add KV for your context length in context VRAM. Overall fit: VRAM fit.
Educational. MoE active params may differ from total params.