Google AI Studio has been around for a while.

We covered it when Gemini launched, and it's been one of the better tools for quickly prototyping AI-powered apps without writing much code.

Last week, Google overhauled it significantly.

The new version is a full-stack development environment i.e databases, user authentication, real-world API integrations, all accessible from a single browser tab.

They're clearly planting their flag in the vibe coding space, and the upgrade is substantial enough to be worth a proper look.

But before that, let’s catch up on AI this week:

Before we get into today's edition, if you're a professional who wants to go from knowing about AI to building with it, this weekend is a good place to start.

📅 28th March | 🕰 10AM to 7PM IST / ET

Session 1 | Getting Started with Generative AI
Session 2 | Building Personalised AI Agents

📅 29th March | 🕰 10AM to 7PM IST / ET

Session 3 | Building Products Using AI
Session 4 | Visual Storytelling & Content Creation Using AI
Session 5 | Mastermind Graduation

It's free. 25 spots left!

1. Dimension: A proactive AI assistant for engineering teams that pulls context from your calendar, email, Slack, and Drive to take action before you ask. Unlike typical AI tools that wait for prompts, it's designed to cut the 30% of engineering time typically spent on busywork rather than shipping.

2. Moda: Launched this week with a $7.5M seed round led by General Catalyst, Moda is a brand-aligned design agent that generates fully editable presentations, social posts, and marketing assets on a layered canvas.

3. Omma by Spline: Spline's new AI canvas takes natural language prompts and produces interactive, production-ready web experiences in a single workflow, skipping the prototype-to-developer handoff entirely. It runs multiple AI agents in parallel and can ingest data inputs like CSV, JSON, and 3D model files to make the outputs dynamic.

It's Google's free, browser-based tool for building apps using natural language.

You describe what you want, and a coding agent called Antigravity builds it, including frontend, backend, and everything in between.

It does not require any downloads, or a setup, and no prior coding experience needed.

It runs on Gemini 3.1 Pro, which is also free to access through AI Studio for now.

The new version is built around four capabilities that didn't exist (or didn't work reliably) before. Here's what each one looks like in practice.

Real-time collaboration tools

Prompt:

Brainstorm Board

What this tests: real-time multiplayer sync, WebSocket connections, live state across users.

How it went: Took two prompts to get here. The first build worked but missed the room code flow and users couldn't actually join each other's sessions.

A follow-up prompt fixed it.

The agent takes your brief literally, so if something isn't explicitly in the prompt, don't expect it to assume.

Real-world data integrations

Prompt:

CityScout

What this tests: Google Maps API integration, Secrets Manager for API keys, Firestore for persistent search history.

How it went: First run threw errors as the app needs a Google Maps API key and I hadn't added it yet.

Once I dropped the key into the Secrets Manager and refreshed, everything resolved.

The game itself was spot-on: room codes worked, the leaderboard updated in real time, the timer ran cleanly.

Games and interactive experiences

Prompt:

Neon Trivia

What this tests: multiplayer rooms, turn-based logic, live leaderboard.

How it went: Same situation as the trivia game i.e. missing API key upfront, errors on first load.

Added the key to Secrets Manager, refreshed, and the app sorted itself out completely.

The Maps integration, the color-coded pins, the search history and all of it worked without any additional prompting.

One thing worth knowing before you open the tool: the prompts that produce good outputs are more detailed than you'd expect.

The specificity helps the agent make the right decisions upfront rather than guessing.

And the prompts above are a reasonable template for it.

Here's the basic flow:

Step 1: Go to aistudio.google.com/apps

Step 2: Write your prompt in the Build tab

Describe your app in plain language. Include: what it does, key features, visual style, and any external services it needs to connect to.

Step 3: Let the agent build

Antigravity will start generating your app in real time. You'll see the code and a live preview side by side.

Step 4: Enable Firebase when prompted

If your app needs a database or user login, the agent will detect this and ask you to enable Firebase.

Click "Enable Firebase" and it handles the rest.

Step 5: Add API keys if needed

For apps that connect to external services like Google Maps, go to Settings and add your credentials in the Secrets Manager. The agent picks them up automatically.

Step 6: Share or continue building

Your app gets a live URL you can share immediately. Come back later and it picks up where you left off.

I've been using AI Studio for a while, mostly to mock up ideas quickly and share them with the team.

If the idea needed a database, or login, or any real functionality, I'd hit a wall and hand it off.

What this update does instead is remove the ceiling.

Gemini isn't winning the public sentiment right now, but I think people are underrating what Google actually has here, which is the full stack.

Lovable and the others are either building that infrastructure from scratch or partnering with Supabase to fill the gaps.

Google already has the parts.

I'd pick AI Studio over other vibe coding tools right now.

The integrations are there, and the tool keeps getting better.

Use one of the prompts above as a starting point, swap in your own idea, and see how far it gets.

Reply with the links to your builds. I may feature my favourite next week.

Until next time,
Vaibhav 🤝🏻

If you read till here, you might find this interesting

#AD 1

Every year, $240B is spent on treating the symptoms of osteoarthritis. But not a single therapy has been able to actually stop it. The answer, it turns out, has been inside us all along.

A startup named Cytonics discovered the human body already produces a protein designed to protect cartilage. It just doesn’t produce enough where it's needed most. So Cytonics harnessed it. 

Their first-generation therapy has already treated 10,000+ patients. Now they've engineered a 200% more potent, mass-producible version pushing toward FDA approval.

If approved, it could be the first therapy to actually halt cartilage destruction and promote regrowth in a market that has never had a real solution. Claim a piece at the pre-clinical stage as an early-stage investor before March 26 to receive time-sensitive investor bonuses.

Invest in Cytonics Now

^{This is a paid advertisement for Cytonics Regulation CF offering. Please read the offering circular at https://cytonics.com/}

#AD 2

This guide distills 10 AI strategies from industry leaders that are transforming marketing.

Outpace your competitors by mastering AI.

Get Your Free Guide