community pool / model answer

SuperDeepseek-V4-Flash-abliterated-MQ-2xDGX

chat · Jiunsong/SuperDeepseek-V4-Flash-abliterated-MQ-2xDGX

68.3 tok/s · model median · 2 runs in the model record.

Reference rigQuantizationMedian resultBasisTTFT (ms)Peak VRAMMax contextEngines
GB10 Grace Blackwell 256GB4-bit37.3 tok/sreported-- GB2,048vllm
GB10 Grace Blackwell 256GB8-bit36.3 tok/sreported-- 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.