Qwen3-Coder 30B A3B

Q4_K_M running on RX 6950 XT

PERFORMANCE / TOKENS PER SECONDRDNA 2

Prompt processing587.3tok/s

#8 in index

Token generation47.8tok/s

#13 in index

Run fingerprint

Enough context to make the number useful.

ModelQwen3-Coder 30B A3BMoE architecture

QuantizationQ4_K_MExact submitted variant

GPURX 6950 XT16 GB VRAM

GPU architectureRDNA 2Metal backend

Test context512prompt tokens

Model familyMoEncmoe 23

Prompt processing and generation are reported separately so memory bandwidth and decode behavior stay visible.

Quantization, GPU family, VRAM and context travel with the result instead of being hidden behind one score.

This result remains explicitly labeled as a community submission.