community pool / model answer

Qwen3.8-Flash-Next-FP8

chat · Qwen/Qwen3.8-Flash-Next-FP8

32.4 tok/s · model median · 12 runs in the model record.

Reference rigQuantizationMedian resultBasisTTFT (ms)Peak VRAMMax contextEngines
Arc Pro B70 32GB ×48-bit32.4 tok/smeasured549- GB4,352vllm

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.