Nowadays, it's no longer a problem to have generative AI create a text, an image or even a video. Why not a finished website? That's exactly what GPT Engineer was built for. After logging in, you describe what kind of website you have in mind in a text prompt and the AI takes care of the rest.
This worked without any major problems with our small text. Changes to individual objects or the entire website could also be made without any problems via the chat interface.
All changes are displayed to the right of the chat area. The chatbot itself divides the respective task into several sub-steps before executing it. If you wish, you can also view changes in a source code view. Here you can see the new features directly next to the old code.
Also practical: If you have GPT Engineer create a new website for you, you can also give the chatbot a screenshot. The AI then tries to base the design on the template. The result is not always perfect, but basic design specifications can be quickly transmitted to the tool.
Nothing behind it? GPT Engineer can do more than “just” frontend development
Basically, GPT Engineer currently only creates the frontend of your website – including the part visible to users. For a simple web presence that only provides static information, this would be perfectly sufficient. But what if you want to build a modern website that processes user input, for example?
Editor's recommendations
For this case, GPT Engineer offers a Supase integration. Supabase is a backend-as-a-service that provides you with a database or user authentication via the cloud. Once connected and set up, you can use this method in GPT Engineer to set up a login mechanism for your users via a prompt.
At this point, it becomes clear that the target group of the tool is probably not the restaurant owner without web know-how who just wants to build a simple web presence. They would still be well advised to use a classic homepage builder.
Rather, it is likely to appeal to small teams and freelancers who have a certain level of knowledge about web technologies and, above all, are looking for a way to become more productive.
This is also reflected in the GitHub integration of GPT Engineer. Once set up, the code generated by the AI is stored in a repository for you and can be processed directly from there. If you make changes to the code, these are then also reflected in GPT Engineer.
This is how much the AI tool costs you
If you just want to try out GPT Engineer, you can do so for free after registering. However, you will then only have a limited number of AI changes available to you per day. In addition, all of your projects are public. For $19 per month, you can remove these restrictions.