community pool / model answer

microsoft_Phi-4-mini-instruct-GGUF

chat · bartowski/microsoft_Phi-4-mini-instruct-GGUF

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

Reference rigQuantizationMedian resultBasisTTFT (ms)Peak VRAMMax contextEngines
Arc Pro B70 32GB4-bit96.6 tok/sreported70- GB4,096llama.cpp
Arc Pro B70 32GB8-bit72.3 tok/sreported119- GB4,096llama.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.