google

gemma-4-26B-A4B-it — performance reports

Community-submitted & scraped throughput measurements. Sortable by generation speed.

← Back to model
Ran this model? Log in to submit your own performance result.
Rig / GPU Quant Runtime Context Prompt tok/s Gen tok/s VRAM (GB) Source Reporter
RTX 3090 GGUF · Q4_K_M llama.cpp 4,096 3,625.60 119.40 — scraped anonymous
RTX 3090 GGUF · Q4_K_M llama.cpp 16,384 3,068.90 115.00 — scraped anonymous
RTX 3090 GGUF · Q4_K_M llama.cpp 32,768 2,453.40 107.50 — scraped anonymous
RTX 3090 GGUF · Q4_K_M llama.cpp 65,536 1,765.10 98.90 — scraped anonymous
RTX 3090 GGUF · Q4_K_M llama.cpp 131,072 1,147.10 83.00 — scraped anonymous
RTX 3090 GGUF · Q4_K_M llama.cpp 262,144 671.40 64.40 — scraped anonymous

selected to compare · pick at least 2

Compare ()