community pool / model answer

Qwen2.5-0.5B-Instruct

chat · Qwen/Qwen2.5-0.5B-Instruct

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

Reference rigQuantizationMedian resultBasisTTFT (ms)Peak VRAMMax contextEngines
AMD Ryzen 7 7840HS4-bit128.3 tok/sreported-- GB2,048llama.cpp
Radeon AI Pro R9700 57GB8-bit102.9 tok/sreported129- GB2,048llama.cpp
RTX 3060 12GB8-bit396.2 tok/sreported2012.4 GB4,608llama.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.