community pool / model answer
Qwen3.8-27B-W4A16-AutoRound-fast
chat · born2bewild/Qwen3.8-27B-W4A16-AutoRound-fast
114.8 tok/s · model median · 54 runs in the model record.
| Reference rig | Quantization | Median result | Basis | TTFT (ms) | Peak VRAM | Max context | Engines |
|---|---|---|---|---|---|---|---|
| CMP 170HX 64GB | 4-bit | 64 tok/s | measured | 11,802 | - GB | 131,336 | sglang, vllm |
Context tested
The table above is the tested context for each exact rig and quantization cell. Empty fields remain empty rather than being filled with an estimate.