LLM Reference

Gemma 4

Google DeepMindApache 2.0
8 models2026Up to 256K ctx

About

Google's most capable open-source model family, purpose-built for advanced reasoning and agentic workflows. Delivered in four sizes (E2B, E4B, 26B MoE, 31B dense) with multimodal capabilities including text, image, video, and audio processing for the smaller models.

Specifications(8 models)

Gemma 4 model specifications comparison
ModelReleasedContextParametersMultimodalFn Calling
Gemma 4 E2B2026-04128k2000000000YesYes
Gemma 4 E2B IT2026-04128k2000000000YesYes
Gemma 4 E4B2026-04128k4000000000YesYes
Gemma 4 E4B IT2026-04128k4000000000YesYes
Gemma 4 26B A4B2026-04256k26000000000YesYes
Gemma 4 26B A4B IT2026-04256k26000000000YesYes
Gemma 4 31B2026-04256k31000000000YesYes
Gemma 4 31B IT2026-04256k31000000000YesYes

Frequently Asked Questions

What is Gemma 4?
Google's most capable open-source model family, purpose-built for advanced reasoning and agentic workflows. Delivered in four sizes (E2B, E4B, 26B MoE, 31B dense) with multimodal capabilities including text, image, video, and audio processing for the smaller models.
How many models are in the Gemma 4 family?
The Gemma 4 family contains 8 models.
What is the latest Gemma 4 model?
The latest model is Gemma 4 E2B, released in 2026-04.

Models(8)