Gemma 4: Google's Apache 2.0 Open-Weight Models for Local Hardware

Melvin Vivas · X post · 2026-04-03 · Open on X

Topics: LLM Fundamentals, Fine-tuning & Model Customization, Industry Trends & Job Market · Level: intermediate

Summary

This post shares Google's Gemma 4 launch. Gemma 4 is a series of open-weight models under the Apache 2.0 license, built to run on phones, laptops and desktops. It comes in a 26B mixture-of-experts (MoE) size and a 31B dense size. The creator asks Unsloth AI when they will support fine-tuning it.

Key points

• Gemma 4 is an open-weight model series released under the permissive Apache 2.0 license.
• Google calls it 'byte for byte' the most capable open model family.
• It is designed to run on your own hardware: phones, laptops and desktops.
• There are two flagship sizes: a 26B mixture-of-experts (MoE) model and a 31B dense model.
• Unsloth AI is the go-to library for quickly fine-tuning new open models like Gemma.

Resources mentioned

• Gemma 4 · tool · ai.google.dev · free
  Google's family of open-weight models in several sizes, built to run on devices and offline, with multimodal and agentic abilities, and open to fine-tuning.
  Also in: Gemma 4 Runs Locally On-Device in the Antigravity SDK (Melvin Vivas on X · notes), On-Device AI: Running Gemma 4 E2B Offline on an iPhone with LiteRT (Melvin Vivas on X · notes), Running Gemma 4 Models Offline on an iPhone (Melvin Vivas on X · notes), Fine-tuning Gemma4-E2B on your own tweet style with Unsloth (Melvin Vivas on X · notes) and 25 more
• Unsloth · tool · unsloth.ai · free
  Open-source library for fast, memory-efficient fine-tuning of open LLMs (LoRA/QLoRA) on a single GPU or in Colab.
  Also in: Run Laya Decision models locally with Unsloth on 4GB RAM (Melvin Vivas on X · notes), Unsloth passes 500M model downloads on Hugging Face (Melvin Vivas on X · notes), Run Qwen-Image-2.1 locally on 12GB VRAM with Unsloth GGUFs (Melvin Vivas on X · notes), Base vs fine-tuned Gemma 4 E2B as a model router (Melvin Vivas on X · notes) and 29 more

Try this

• Watch for Unsloth support so you can fine-tune Gemma 4