Gemma 4

language modelGoogle

Gemma 4 31B Instruct is Google DeepMind's 30.7B dense multimodal model supporting text and image input with text output. Features a 256K token context window, configurable thinking/reasoning mode, native function... License: apache-2.0.

Key facts

Made byGoogle
Typelanguage model
ReleasedMarch 2, 2026
Licenseapache-2.0
Reads at once262,144 tokens
Price (API, per million tokens)$0.08 in · $0.26 out
Input → outputimage, text, video → text
Sizes26B, 31B, 12B
Downloads (recent)583.1K
Likes on Hugging Face820

Questions about Gemma 4

Who makes Gemma 4?

Gemma 4 is made by Google.

When was Gemma 4 released?

We first saw Gemma 4 listed on March 2, 2026.

How much does Gemma 4 cost?

Through OpenRouter it costs $0.08 per million input tokens and $0.26 per million output tokens. Prices change, so check the provider before you build on it.

How much text can Gemma 4 read at once?

Up to 262,144 tokens, roughly 196.6K words.

More from Google

Other language models

Modelcast collects this page automatically from public sources (OpenRouter, Hugging Face). Facts are copied from those sources, not written by the maker. Not affiliated with Google.