community pool / model answer

Qwen3.8-27B-W4A16-AWQ

chat · philbert440/Qwen3.8-27B-W4A16-AWQ

103 tok/s · model median · 6 runs in the model record.

Reference rigQuantizationMedian resultBasisTTFT (ms)Peak VRAMMax contextEngines
RTX 5090 32GB4-bit103 tok/smeasured-- GB2,048vllm

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.