Overview
The high-quality dense Gemma 4 flagship for local reasoning, coding, and agent workflows.
Gemma 4 31B
The high-quality dense Gemma 4 flagship for local reasoning, coding, and agent workflows.
MODEL NUMBER105
- Release date
- 2026-04-02
- Organization
- Google DeepMind
- Family
- Gemma
01
Specifications
Only published specifications are shown.
02
Overview
PreviousGemma 3 27B27B · 128K ctx
CurrentGemma 4 31B31B · 256K ctx
New capabilities Reasoning Tool use Coding
04
Architecture & capabilities
Key capabilities
ReasoningTool useCodingMultilingual
05
Release history
Gemma 4 31B launched.
View source06
Related models
Select a card to view the related model.
07