Gemma 4
Google DeepMind's open-weight multimodal AI model family — run locally or in the cloud
About Gemma 4
Overview
Gemma 4 is Google DeepMind's latest generation of open-weight AI models, available in four sizes (E2B, E4B, 26B, 31B) from mobile-class to high-performance inference. The models support multimodal reasoning (audio and vision), native function calling for agentic workflows, and 140+ languages — with the flagship 31B variant competing with frontier proprietary models on benchmarks.
Key Features
- Completely free and open-weight — no API costs, full local control
- Four model sizes covering mobile/edge to high-performance use cases
- Multimodal out of the box — audio and vision without extra models
- Huge ecosystem: Hugging Face, Ollama, LM Studio, Docker, Vertex AI, GKE
- Native function calling built into the architecture
Open-source workflow automation for technical users.
AI-powered website builder for professional websites
AiWind prompt library for Midjourney and AI image prompts
The viral AI platform for chatting with any character.
Community platform for discovering models and creating AI art
Kanban app for orchestrating coding agents across isolated worktrees