Gemma 4 E4B

Gemma 4 E4B is Google DeepMind's compact multimodal model with 4.5 billion effective parameters (8B with embeddings) and a 128K context window. Supports image, text, and audio inputs.

MMMLU

76.6%

i
MMLU-Pro

69.4%

i
MathVision

59.5%

i
GPQA

58.6%

i
t2-bench

57.5%

i
MMMU-Pro

52.6%

i
LiveCodeBench v6

52.0%

i
AIME 2026

42.5%

i
BIG-Bench Extra Hard

33.1%

i
MedXpertQA

28.7%

i
MRCR v2

25.4%

i