July 2026

Modern AI systems such as ChatGPT, Codex, and Claude have made developing custom applications much easier. Today, almost anyone can create an individual control application for vMix within a short period of time, even without extensive programming knowledge.

It is no longer essential to know how to write every line of VB.NET, C#, or JavaScript yourself. What matters much more is being able to clearly describe to the AI what the application should do.

For example, a task could be described like this:

Create a small VB.NET application that uses the vMix HTTP API
to switch Input 3 to Program and display Input 5 as an overlay.

Use a central function for sending the commands and display
an error message if vMix cannot be reached. 

Based on such a description, a modern AI system can generate the required code, the user interface, and the basic project structure. Further changes can also be described in normal language.

For example:

Add a second button that hides the overlay again.

Development is often even faster and more structured when existing projects or code examples are used as templates.

My GitHub profile contains several examples for controlling vMix with VB.NET:

GitHub profile peteraellig

These include examples for:

* controlling vMix through the HTTP API * controlling vMix through the TCP API * standardised handling of vMix commands * title and graphics control * sports displays and game clocks * external control applications * GPIO connections

These projects can be referenced when working with an AI system or used directly as a starting point.

A suitable instruction could be:

Use the GitHub project peteraellig/vMix-Control as a template.

Use the existing function for sending vMix commands and extend
the application with (your idea) 

This means the AI does not have to develop every function from scratch. Existing structures, methods, and practical experience can be reused and adapted for a new task.

A short description is often enough to begin creating a custom application:

* What should be controlled? * Which inputs or titles are used? * Which buttons are required? * In which order should the actions be carried out? * Should the application run on the same computer or communicate over a network?

Based on this information, an AI system can create a suitable solution step by step.

Custom control applications can be created for tasks such as:

* switching cameras * showing and hiding overlays * changing titles and text fields * starting recordings or streams * controlling sports displays * executing several commands with a single button * creating an interface designed for a specific production

The application does not need to be complete from the beginning. You can start with one simple function and then expand it step by step together with the AI.

Even modern AI systems do not always produce error-free code immediately. Common causes include unclear instructions, incorrect input names, or differences between vMix versions.

It is helpful to provide the AI with the following information:

* the vMix version being used * the preferred programming language * existing code * exact error messages * the required workflow

Generated code should always be checked before use. If an error occurs, the complete error message can be passed back to the AI. In my experience, the AI can correct the code directly in almost every case.

With the help of a modern AI system, almost anyone can now develop a custom control application for vMix within a short period of time.

You describe the required function in normal language, have an initial application created, and then expand it step by step.

Existing examples, such as those available on my GitHub profile, can be used as templates. This often results in better-structured code and speeds up the development process.

In this way, an idea for a specific project can quickly become an individually adapted vMix application.