---
name: Gemma 4 12B Instruct
slug: gemma-4-12b
type: Multimodal
providers:
  - nano-gpt
description: >-
  Google's Gemma 4 12B Instruct is an open-weight multimodal model for text,
  image, audio, and video understanding, with tool calling and structured output
  support.
tags:
  - open-source
  - gemma
  - multimodal
  - tool-use
is_open_source: true
context_window: 262144
max_output: 32768
pricing:
  - image: 0
    input: 0.06
    output: 0.3
    provider: nano-gpt
release_date: '2026-08-01'
capabilities:
  audio: true
  vision: true
  tool_use: true
  citations: false
  reasoning: false
---
Gemma 4 12B Instruct is a Multimodal model.
Google's Gemma 4 12B Instruct is an open-weight multimodal model for text, image, audio, and video understanding, with tool calling and structured output support.