community pool / model answer

MiMo-V2.5

chat · XiaomiMiMo/MiMo-V2.5

32.8 tok/s · model median · 5 runs in the model record.

Reference rigQuantizationMedian resultBasisTTFT (ms)Peak VRAMMax contextEngines
Radeon AI Pro R9700 32GB ×32-bit33.2 tok/smeasured11,694- GB32,768llama.cpp
Ryzen AI Max 395 128GB2-bit28.1 tok/sreported3,919- GB768llama.cpp

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.