GCP – Create shareable generative AI apps in less than 60 seconds with Vertex AI and Cloud Run

Want to turn your generative AI ideas into real web applications with one click? 

Any developer knows it’s a complex process to build shareable, interactive applications: you have to set up infrastructure, wire APIs, and build a front-end. It’s usually a complex process. What if you could skip the heavy lifting and turn your generative AI concept into a working web app with just a few clicks?

Today, we’re thrilled to introduce a streamlined workflow within Google Cloud’s Vertex AI: one-click deployment of your generative AI prompts directly to Cloud Run as interactive web applications. We’ll show you how it works, and how you can get started.

Bridging the gap from prompt to prototype

Vertex AI provides a helpful environment for experimenting with and refining generative AI prompts. You can test different models, tune parameters, and craft the perfect instructions. However, sharing that interactive experience beyond the console often means exporting code, setting up hosting, managing dependencies, and building a user interface.

Finding the right path to deployment isn’t always straightforward. The missing piece – a clear option to transform the prompt into a shareable prototype.

The solution: Simple, fast deployment with “Deploy as App”

Based on this feedback and our goal to make generative AI accessible, we’ve integrated a seamless deployment path:

1. Craft your prompt: Perfect your generative AI idea within the familiar Vertex AI Studio interface. Add system instructions, examples, and test until you’re happy.

2. Click “Deploy as App”: We’ve introduced a clear, primary “Deploy as App” button. No more ambiguity – this is your direct path to creating an application.

3. Configure as needed: Select your authentication preference (public or authenticated) as you choose.

4. Vertex AI and Cloud Run do the heavy lifting: Click “Deploy application,” and Vertex AI works with Cloud Run behind the scenes. Vertex AI packages your prompt and builds the interactive UI (powered by Gradio), while Cloud Run handles building and hosting the app in its fully managed infrastructure. You get real-time status updates along the way.  

5. Share your app. Once complete, you get a direct link to your live, functional web application powered by your prompt. You can easily share this URL with colleagues, stakeholders, or testers.

6. Seamlessly iterate: Easily return to Vertex AI Studio, refine your prompt, and redeploy the application with your changes.

Customize with Cloud Run

Your app is deployed to Cloud Run, Google Cloud’s fully-managed application platform. This means you get automatic scaling (including to zero for cost savings) without managing infrastructure – perfect for quickly deploying and sharing your apps.  . To customize your app and take it to production, you can edit the application code directly in Cloud Run’s source editor to make it your own. You can also download the code, and use your IDE of choice, pushing updates with Cloud Run’s git integration.

Why this matters:

• Speed: Go from concept to a shareable proof-of-concept in minutes, not hours or days.

• Simplicity: Focus on your AI prompt and idea, not on complex infrastructure setup.

• Iteration: Easily return to Vertex AI Studio, refine your prompt, and redeploy the application with your changes.

• Shareability: Instantly get a working web interface to demonstrate your generative AI’s capabilities.

The Cloud Run integration lowers the barrier to entry for creating and sharing generative AI applications. Whether you’re building a quick prototype, an internal tool, or a demo for stakeholders, Vertex AI Studio now provides an incredibly efficient path from prompt engineering to a live, interactive experience powered by Cloud Run.

Get started

Ready to bring your generative AI ideas to life? Head over to Vertex AI Studio in the Google Cloud console and look for the “Deploy as App” button. We can’t wait to see what you build!