Hey there,

Just last month, Google rolled out their own vibe coding tool, Build Mode. This week they gave it a massive update, and after testing the features, AI Studio has become my new favorite vibe coding tool. Let’s break down what’s changed. 

How did Build Mode Change?

Google has significantly expanded what you can create within Build Mode. From simplifying the creation process with building blocks to increasing customization with the annotation tool, Build Mode looks very different than it did last month. 

Annotation Mode: 

We’ve all felt the frustration of wanting to change something using AI and either struggled to describe exactly what we want or simply wanted to speed up the process. This is where the Annotate tool comes in.

AI Studio now allows you to draw directly on your screen to better explain the changes you’re looking for. In the example below, I’m able to explain that the base of my game is off screen and needs to be moved to where I indicated. This is a game-changer for vibe coding and for improving the iterative development process. It cuts out wasted time and gives you greater control over how your app is created. 

Building Blocks: 

You can now bring pre-built, AI-powered components directly into your apps with just a few clicks. The building blocks include sixteen unique features that range from image editing with Nanobanana to generating speech. These building blocks allow users of all expertise levels to easily add powerful features to their apps. It also means that you don’t need to worry about manually connecting your apps to the Gemini API. Ultimately it gives more control to developers. 

Improved Brainstorming: 

AI Studio has added several new features to help with the initial brainstorming process. Users can now use features like the “Feeling lucky” button to auto-generate an app idea. This seems more like a fun feature to showcase what build mode can do but isn’t actually practical. The “Suggestions” list however, brainstorms feature ideas in realtime. It gives great ideas and advice to help elevate your creations. Again this new feature seems to improve the ease-of-use of Build Mode and makes the platform feel more polished. 

Build Mode in Action

The feature list is impressive but the real test is to see what can be created. I created two very different apps using AI Studio’s new tools and got some very interesting results. 

Demo 1: A Retro Arcade

• The Idea: In this example, I asked Gemini to create a virtual arcade filled with classic games. Things like Tetris, Pong, and Galaga. 

• Tools Used: While I didn’t use any building blocks for this project, I did use the annotation tool to fix a persistent bug I had. I also used the “Suggests” to add things like custom audio to each game. 

• The Result: I think this early stage app was a success. I’ve only spent about 30 minutes on this project and created a solid foundation for further development. 

Demo 2: Research Assistant

• The Idea: For my second example, I had Build Mode create a research assistant that could access Google search results and read websites. This was meant to be a flexible, AI-powered search assistant. 

• Tools Used: I was able to use the Google Search and Thinking Mode building blocks for this app. This helped me quickly connect my app to the Gemini API without needing to write any code. 

• The Result: This app was actually a one-shot prompt. The core logic and functionality were generated perfectly on the first try. Now the powerful backend is ready for some UI refinements.

The AI Studio team has created a massive catalogue of apps to demonstrate the full capacity of the updated Build Mode. You can explore and customize these apps for your own use. The Gallery has been overhauled and significantly expanded to include over forty unique apps. 

My Thoughts

Overall I think that Google has made Build Mode more accessible for non-developers and accelerated the creation process for experienced developers with their latest updates. I’ve explored a number of other vibe coding tools including Bolt and Lovable and feel that AI Studio’s updated Build Mode out performs them. 

• For Non-Developers: The latest updates seem to be targeted at non-developers, giving them the tools to create powerful, technical apps without needing to write code. The new tools allow a user to go from a vague idea (or no idea) to a basic AI-powered creation in seconds. They can then use the suggestions to make further improvements without needing to even open the code view. 

• For Developers: These updates are still very helpful. Now you can add complex AI features with the click of a button instead of manually coding it and have Gemini do the heavy lifting while you focus on any potential errors or complex additions. You can also use the smaller updates like “Save to GitHub”. 

The greatest advantage to Build Mode in my eyes is the fact that it's fast, free, and powerful. While Build Mode was always free for people with Google accounts, I think it's worth reiterating here that you can access all of these impressive features for free. While I couldn’t find the exact rate limits, I was able to spend hours in Build Mode without hitting any limits. Currently the only paywalled feature is using Veo 3 to generate videos. 

While AI Studio has cemented itself as my go-to vibe coding tool, I’d suggest you experiment with it to see if it is a good fit for you.