ammaarreshi

ammaarreshi / gemma-chat

Public

Local AI chat + coding agent for Apple Silicon, powered by Gemma 4 via MLX / Supports Ollama

29
5
100% credibility
Found Apr 29, 2026 at 29 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
TypeScript
AI Summary

Gemma Chat is a Mac desktop app that runs Google's Gemma AI models locally for conversational assistance and generating code projects with real-time previews entirely offline after initial setup.

How It Works

1
🔍 Discover Gemma Chat

You stumble upon this offline AI buddy for chatting and coding right on your Mac while browsing GitHub.

2
💻 Launch the app

Grab the app, start it up, and it handles the one-time setup by downloading what it needs.

3
🤖 Choose your AI size

Pick a model that fits your Mac – small and fast or bigger and smarter – and watch it get ready.

4
Pick chat or build
💬
Chat casually

Ask questions, get ideas, or brainstorm like with a friend.

🔨
Build a project

Describe an app or page, and see it come together step by step.

5
🎤 Tell it what you want

Type or speak your idea, like 'make a fun timer' or 'explain this topic'.

6
✨ Watch the magic happen

The AI types code live, files appear, and a preview shows your creation updating in real time.

🎉 Your offline win

Enjoy your new app, answers, or project – all working anywhere, no internet needed.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 29 to 29 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is gemma-chat?

Gemma-chat is a local ChatGPT alternative and Gemma chatbot that runs entirely offline on Apple Silicon Macs, using Google's Gemma 4 models via Apple's MLX framework. Built as a TypeScript Electron app with React and Tailwind, it offers two modes: chat for Q&A with tools like web search and calculator, or build mode where you describe a web app and it generates multi-file code with a live preview canvas. No API keys or internet needed after the initial 3GB model download—perfect for private, unrestricted local chat AI.

Why is it gaining traction?

It stands out as a local ChatGPT model runner with zero-config setup that auto-provisions MLX and supports model switching from 1.5GB lightweight to 18GB beasts, plus voice input via browser Whisper. Developers dig the offline coding agent that streams file writes in real-time, letting you iterate on HTML/CSS/JS projects without cloud leaks. Voice-to-text and sandboxed workspaces make it a seamless local chat server for quick prototypes.

Who should use this?

Frontend devs prototyping landing pages or games on flights, indie hackers building MVPs without subscriptions, and AI tinkerers wanting a local chatbot AI for code gen or research. Ideal for Apple Silicon users needing a Gemma chat template alternative to cloud tools, especially those running local GitHub Actions in isolated workspaces.

Verdict

Grab it if you're on M-series Mac and want offline Gemma-powered coding—solid POC despite 29 stars and 1.0% credibility score. Early maturity means watch for bugs, but MIT license and clear README make it easy to fork and improve.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.