Gemma 4 26B-A4B

UD-Q4_K_M running on AMD Radeon Pro Vega II + AMD Radeon PRO W6800X

PERFORMANCE / TOKENS PER SECONDGCN / Vega + RDNA 2

Prompt processing484.84tok/s

#23 in index

Token generation48.18tok/s

#25 in index

Run fingerprint

Enough context to make the number useful.

ModelGemma 4 26B-A4BMoE architecture

QuantizationUD-Q4_K_MExact submitted variant

GPUAMD Radeon Pro Vega II + AMD Radeon PRO W6800X63.97 GB VRAM

GPU architectureGCN / Vega + RDNA 2Metal backend

Test context512prompt tokens

Model familyMoESigned ToshLLM app · amd-gpu · Metal

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.