Bring state-of-the-art agentic skills to the edge with Gemma 4

Source: Google Developers Blog

Overview

Google DeepMind has launched Gemma 4, a family of state‑of‑the‑art open models designed to enable multi‑step planning and autonomous agentic workflows directly on‑device. The release includes:

• Google AI Edge Gallery – a platform for experimenting with “Agent Skills”.
• LiteRT‑LM library – provides a significant speed boost and structured output for developers.

Gemma 4 is available under an Apache 2.0 license, supports over 140 languages, and is compatible with a wide range of hardware, including mobile devices, desktops, and IoT platforms such as Raspberry Pi.