community pool / model answer

Qwen3-8B-GGUF

chat · unsloth/Qwen3-8B-GGUF

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

Reference rigQuantizationMedian resultBasisTTFT (ms)Peak VRAMMax contextEngines
Ryzen AI Max 395 128GB4-bit39.3 tok/sreported996.2 GB4,096llama.cpp
Ryzen AI Max 395 64GB4-bit40.9 tok/sreported1,1316.1 GB8,192llama.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.