community pool / model answer

gemma-4-E4B-it-qat-q4_0-gguf

chat · google/gemma-4-E4B-it-qat-q4_0-gguf

84 tok/s · model median · 4 runs in the model record.

Reference rig	Quantization	Median result	Basis	TTFT (ms)	Peak VRAM	Max context	Engines
RTX 3060 12GB	4-bit	88.1 tok/s	reported	1,371	4.9 GB	4,608	llama.cpp
RTX 5070 12GB	4-bit	152.7 tok/s	reported	60	- GB	32,768	llama.cpp
RX 570 4GB	4-bit	27.5 tok/s	reported	326	- GB	2,048	llama.cpp
RX 570 4GB ×2	4-bit	33.2 tok/s	reported	242	- GB	2,048	llama.cpp

Context tested

The table above is the tested context for each exact rig and quantization cell. Empty fields remain empty rather than being filled with an estimate.