It is clear that AI will allow us to do many things that we have never doneOne of them is programming. With basic knowledge, it is possible to take advantage of the full potential of Gemini and Google tools that start from it: its hidden “Photoshop” allows you to modify the images, and its integration into the Application Messages A spam scanner. Now I test one of the revealed during the I / or 2025 Google.
I have no idea of the programming, even less interface design. That I would like, there is a lot to say, and that is why Google Stitch called me from the start. Will this application replace the user interface designers? I doubt it, but it is A good start For another AI application with great sense. I tell you my experience.
It's Google Stitch
Stitch is a more Google tool, but not. Like others like the book note, Stitch begins by offering us the possibility of creating a project of Software designbasis of what it allows to do.
All my tests, I carried out with the predetermined mode or the “standard mode” used by Gemini Flash 2.5, the language model focused on response speed. Alternately, The experimental mode understands better the demand that I makeKey that Gemini Pro 2.5 is under the hood. They have a difference that I will specify then, although the rules of the game do not change too much.
What allows you to create? User interfaces, whether for a website or for a mobile application. With Zero Design Knowledge, he generated visually attractive results, very in accordance with what Google himself considers that it should be Android: material 3 bet on colors, also for minimalism.
A simple prompt was sufficient for me to have its magic: “an application to manage tasks”. In a few seconds, there was “what was just a diffuse image in my mind. Fortunately for the most understood, Google allows you to modify each result, in addition to exporting it to Figma. And do it conversely: adding to the rapid code already designed, for example in JSON.
My task application, or rather its design
I imagine that the implementation would take a lot of work behind, but the truth is that it can be useful in a first phase of development. Starting from a sketch or an idea is more effective and in these sewing points, it makes it great. Any modification that I propose it without too much difficulty. Then he leaves me at hand HTML HTML or CSS codePerfect for implementing it in an IDE.
Of course, this is not perfect: the editing options (without adding them in invites) They are rare. As soon as I change the color palette, the application mode between light or darkness and typography. And I remember, simply using the Flash Gemini 2.5 model, otherwise they must pass yes or yes for a next prompt.
Stitch generates code or export fronted on Figma. This is a first step for new projects
Sometimes, AI locks and you can put an ankleIn addition to the occasional delay in the impression of the petition. He shows that he is still developing software, according to Google, it's beta. Now, to start a project and have a first image is not bad.
An example of application to find the paradise islands
As with other similar tools, see Firebase Studio, We are alone at the start. It is appreciated in operation, but everything is about to change. In addition, it is free and does not require a Google AI Pro subscription, nor the most “ultra” face. There is no reason not to give it a chance, if only to imagine as in my case. It is as simple as to follow this link.
In Xataka Android | The Google Gemini Revolution begins with gems. What are they and how to create yours
In Xataka Android | Google has a destination for Gemini: Become the Total Superapp for Android
