community pool / model answer

Qwen3-30B-A3B-GGUF

chat · unsloth/Qwen3-30B-A3B-GGUF

30.7 tok/s · model median · 6 runs in the model record.

Reference rigQuantizationMedian resultBasisTTFT (ms)Peak VRAMMax contextEngines
Radeon 890M 32GB4-bit30.7 tok/sreported1,340- GB2,048llama.cpp
Ryzen AI Max 385 24GB4-bit30.7 tok/smeasured1,340- GB2,048llama.cpp
Ryzen AI Max 395 64GB4-bit89 tok/sreported82117.8 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.