Gemma 4 now features Multi-Token Prediction (MTP) drafters, enhancing inference speed and efficiency. This update aims to improve performance across developer workstations, mobile devices, and cloud environments.

Google's Gemma 4 model family now runs natively on iPhones, enabling full local AI inference offline. Early benchmarks show the 31B variant of Gemma 4 performing comparably to Qwen 3.5's 27B model.

Gemma 4 was tested as a local model in Codex CLI to determine its viability as a replacement for cloud models in coding tasks. The testing involved using a 24 GB M4 Pro MacBook Pro to evaluate Gemma 4's local tool calling capabilities.

Ollama can be installed on a Mac mini with Apple Silicon using Homebrew with the command `brew install --cask ollama-app`, which includes auto-updates and the MLX backend. A minimum of 16GB of unified memory is required for running Gemma 4, and the Ollama app will appear in the Applications folder and the menu bar after installation.

Gemma 4 delivers maximum compute and memory efficiency for mobile and IoT devices, enhancing intelligence in these platforms. It supports the development of autonomous agents capable of planning, navigating apps, and completing tasks, while also offering strong audio and visual understanding for rich multimodal applications and multilingual experiences that extend beyond simple translation.