community pool / model answer

Qwen3.8-Flash-Next-NVFP4

chat · nvidia/Qwen3.8-Flash-Next-NVFP4

112.8 tok/s · model median · 3 runs in the model record.

Reference rigQuantizationMedian resultBasisTTFT (ms)Peak VRAMMax contextEngines
GB10 Grace Blackwell 256GB4-bit62.9 tok/sreported332- GB262,144vllm
RTX PRO 6000 Blackwell 96GB ×24-bit83.8 tok/sreported166- GB2,048vllm
RTX PRO 6000 Blackwell Workstation Edition 96GB ×24-bit260.1 tok/sreported450179.1 GB262,144vllm

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.