Community performance summary

Gemma 4 26B-A4B × Vega II + W6800X Duo

6 signed observations in one performance comparison. Open any run below to inspect its exact configuration and provenance.

6 signed runs · 0 verified · 1 tester
Median prompt925.62tok/s · 904.07 to 1425.95
Median generation57.35tok/s · 55.83 to 64.29
Configurations1UD-Q4_K_M
Last observedSep 16GCN / Vega + RDNA 2

What the results show

One clear view.
Every recorded run.

Measurements from comparable hardware and models are presented together. Medians show typical performance, while every signed run remains available below for inspection and sharing.

Quantizations observed1 groups
UD-Q4_K_M6 runs
macOS

macOS 26.6.2 Tahoe

Recorded evidence

Inspect every run.

Open any result to review its exact hardware, model configuration, measurements and submission details.

1 to 8 of 6 configurations

Verified benchmark Community submission
Model / quantHardwareArchitecturePromptGenerationSource
Gemma 4 26B-A4BUD-Q4_K_M · MoE · Signed ToshLLM app · amd-gpu · Metal · comparison
GCN / Vega + RDNA 2
1425.95tok/s
64.29tok/s
threadedSep 16, 2026
Gemma 4 26B-A4BUD-Q4_K_M · MoE · Signed ToshLLM app · amd-gpu · Metal · comparison
GCN / Vega + RDNA 2
904.07tok/s
57.49tok/s
threadedSep 16, 2026
Gemma 4 26B-A4BUD-Q4_K_M · MoE · Signed ToshLLM app · amd-gpu · Metal · comparison
GCN / Vega + RDNA 2
945.34tok/s
57.36tok/s
threadedSep 16, 2026
Gemma 4 26B-A4BUD-Q4_K_M · MoE · Signed ToshLLM app · amd-gpu · Metal · comparison
GCN / Vega + RDNA 2
948.35tok/s
57.33tok/s
threadedSep 16, 2026
Gemma 4 26B-A4BUD-Q4_K_M · MoE · Signed ToshLLM app · amd-gpu · Metal · comparison
GCN / Vega + RDNA 2
905.89tok/s
57.2tok/s
threadedSep 16, 2026
Gemma 4 26B-A4BUD-Q4_K_M · MoE · Signed ToshLLM app · amd-gpu · Metal · comparison
GCN / Vega + RDNA 2
904.81tok/s
55.83tok/s
threadedSep 15, 2026