New Google Gemma 4 Coder Desktop App is INSANE!
The video introduces Gemma Chat, a free desktop app for Apple Silicon Macs that runs Google's Gemma 4 AI model entirely offline without cloud connectivity. The app enables users to build web apps, landing pages, and tools through conversational prompting with live previews. The presenter argues this represents a major shift toward local AI that prioritizes privacy and eliminates subscription costs.
Summary
The video presents Gemma Chat, a free desktop application for Mac that runs Google's Gemma 4 AI model locally on Apple Silicon hardware. The presenter emphasizes that the app requires no internet connection, no cloud subscription, and no login — the AI runs entirely on the user's machine using Apple's MLX framework, which is optimized for M1 through M4 chips, delivering fast, responsive performance comparable to cloud-based tools like ChatGPT.
A central feature highlighted is the app's ability to generate working web apps, landing pages, forms, quizzes, and tools through simple conversational prompts, with live previews displayed alongside the chat interface. The presenter describes this as 'vibe coding' — a workflow where users describe what they want in plain language and the AI builds it without requiring any coding knowledge. This is contrasted with the traditional approach of hiring developers, which cost hundreds of dollars and days of back-and-forth.
The presenter places strong emphasis on privacy as a key differentiator of local AI. He argues that cloud AI tools log, store, and potentially use everything typed into them, making them unsuitable for sensitive business information like client lists, contracts, and internal documents. With Gemma Chat, all data stays on the user's machine and is deleted when the user chooses.
Regarding Gemma 4's capabilities, the presenter acknowledges that large cloud models like Claude and GPT-4 still outperform it on complex reasoning and long-context tasks, but argues that Gemma 4 handles roughly 80% of typical business AI use cases — drafting emails, building landing pages, summarizing notes, and generating ideas — well enough to be genuinely useful. He attributes Gemma 4's local viability to its position as an open-source model that balances small size with sufficient intelligence.
The presenter frames Gemma Chat and local AI broadly as the beginning of a larger trend that will eventually challenge the cloud subscription model. He predicts that as local models continue improving rapidly — noting they went from nearly unusable a year ago to building functional web apps today — the rationale for paying monthly cloud subscriptions will weaken. The video closes with repeated promotional mentions of the presenter's paid community, the AI Profit Boardroom, and a free community called the AI Success Lab.
Key Insights
- The presenter claims Gemma Chat uses Apple's MLX framework to run Gemma 4 at speeds comparable to ChatGPT on M1–M4 Macs, making local AI feel fast rather than sluggish for the first time.
- The presenter argues that Gemma Chat originated from a Google AI Studio contributor and is being shared through Google's official channels, framing it as a deliberate signal that Google wants Gemma 4 running on consumer machines rather than a random side project.
- The presenter states that Gemma 4 occupies a unique 'sweet spot' — small enough to run on a laptop yet capable of agentic tasks where it plans, acts, checks its work, and continues autonomously, unlike models that are either too small to be useful or too large to run locally.
- The presenter argues that one year ago local AI models 'couldn't write a paragraph that made sense,' but now they can build working web apps, and uses this trajectory to suggest the cloud subscription model will become hard to justify once local models close the remaining gap.
- The presenter contends that for roughly 80% of what business owners actually do with AI — drafting emails, building landing pages, summarizing notes, and writing first drafts — Gemma 4 is already sufficient, with cloud models only meaningfully ahead on hard reasoning and massive context tasks.
Topics
Transcript
[0:00] New Google Gemmafour Coder desktop app is insane. Google just dropped a new desktop app called Gemma Chat and it runs Gemmafour right on your Mac. No internet, no cloud, no subscription. You open the app, you type what you want to build and it builds it right there on your laptop offline. This is the new Google Gemmafour Coder desktop app and I think it's going to change how regular people use AI forever. Let me show you why. So here's what Gemma Chat actually does. It's a free desktop app for Mac. It runs Google's brand new Gemmafour model and it lets you build small web [0:30] apps, games, landing pages, simple tools all from a chat…
Full transcript available for MurmurCast members
Sign Up to AccessMore from Julian Goldie SEO
NEW Nvidia Autonomous AI is WILD!! 🤯
Nvidia announced Nemo Clo, a new autonomous AI agent system that operates independently without continuous prompting. Powered by Nemotron 3 Ultra (a 550 billion parameter model), the system is five times faster and cheaper than previous versions, with OpenShell providing secure sandboxed execution.
Laguna XS 2.1: New FREE + Opensource Local AI!
Julian reviews Laguna XS 2.1, a new free open-source local AI coding model from Poolside that performs comparably to Qwen 3.6 and outperforms Claude Haiku on benchmarks. He demonstrates its practical capabilities by building landing pages and functional apps, highlighting its speed, offline functionality, and multiple deployment options through local setup, Claude Code, or OpenRouter's free API.
How to Run Hermes FREE Forever!
The video demonstrates how to run the Hermes AI agent for free using Gemma 4, a local open-source model from Google, with significant speed improvements through MLX optimization. The setup works on Apple Silicon Macs or via free APIs on Open Router, enabling autonomous agents to work offline and privately without subscription costs.
This NEW Chinese AI is INSANE! (FREE + Open Source!)
Long Cap 2.0 is a new open-source Chinese AI model from a food delivery app company that offers 1 million tokens of free context memory, beats GPT-4.5 on SWE bench pro benchmarks, and uses efficient parameter activation to reduce computational overhead while maintaining high performance.
Claude Code is now FREE: Here’s how…
Google's new Gemma 4 model running on Ollama is 90% faster on Apple Silicon, enabling free Claude Code usage locally without token costs. The setup requires three simple steps: downloading Ollama, Gemma 4, and installing into Claude Code, with alternatives available via OpenRouter API for non-Mac users.