I’ve been using v0 extensively to build and maintain a real application, and the latest changes to the code editing workflow have made the experience significantly worse for me. Previously, I could inspect the code, make small changes directly when appropriate, and use the AI for the parts where it actually added value. This made the workflow fast and efficient. With the new experience, I’m increasingly forced to use the chat even for very small and surgical changes. In practice, this has had three negative effects:

• Tasks that used to take a few minutes now take considerably longer.
• Simple changes require unnecessary AI interactions and consume credits.
• Debugging has become much more cumbersome because inspecting, locating, and changing small pieces of code is harder than before. Today I spent roughly 3x more time solving issues that would have been straightforward with the previous workflow. I understand that increasing AI usage may increase credit consumption and potentially revenue per user. But, at least in my case, the effect is likely to be the opposite. I’m willing to pay more for a tool that makes me significantly more productive. I’m not willing to pay more because the product intentionally or unintentionally introduces friction that forces me to use AI for tasks I could previously perform faster myself. v0 was becoming an important part of my development workflow precisely because it combined AI assistance with the ability to work directly with the code. Removing or restricting that flexibility changes the value proposition considerably. If this workflow remains as it is, instead of spending more on v0, I will probably move the project to another development workflow/tool. My suggestion would be simple: keep AI-first editing, but restore direct code editing as an option. Let users decide when AI is useful and when a manual edit is faster. More AI usage should come from making the AI more valuable, not from making manual work harder.