community pool / model answer

gemma-4-E2B-it-GGUF

chat · unsloth/gemma-4-E2B-it-GGUF

90.4 tok/s · model median · 10 runs in the model record.

Reference rig	Quantization	Median result	Basis	TTFT (ms)	Peak VRAM	Max context	Engines
M5 Pro 24GB	8-bit	77.4 tok/s	reported	201	- GB	4,224	llama.cpp
Radeon 890M 32GB	4-bit	35.2 tok/s	reported	640	- GB	2,048	llama.cpp
RTX 3060 12GB	4-bit	113.4 tok/s	reported	-	1.9 GB	32,768	llama.cpp
Ryzen AI Max 385 24GB	4-bit	35.2 tok/s	measured	640	- GB	2,048	llama.cpp
Ryzen AI Max 395 64GB	4-bit	105.6 tok/s	measured	532	3.8 GB	8,192	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.