Gemma 4 A4B 98e V6 Coder It

Q4_K_M running on AMD Radeon GFX10, 16 GB with macOS 15.7.8 Sequoia

PERFORMANCE / TOKENS PER SECONDRDNA

Prompt processing1773.77tok/s

#17 in index

Token generation94.17tok/s

#24 in index

Run fingerprint

Enough context to make the number useful.

ModelGemma 4 A4B 98e V6 Coder ItMoE architecture

QuantizationQ4_K_MExact submitted variant

GPUAMD Radeon GFX10, 16 GB15.98 GB VRAM

GPU architectureRDNAMetal backend

Test context512prompt tokens

Model familyMoESigned ToshLLM app · amd-gpu · Metal

ToshLLM versionv0.83.14App version used to capture and sign this run

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.