Model lifecycle
Learn how your model can develop over time.
An AI Builder model goes through several stages as it evolves from when you create it through publishing, sharing, and use. The following sections describe each of those stages.
Create your model
You can find all the AI Builder model types and capabilities on the AI models page, which you can access from the left menu. This page is where you start to create a new custom AI model, or select a prebuilt model that you can use right away.
Draft models
When you've made progress building and customizing your AI model, you can save a draft to finish later. You can come back at any time and continue where you left off.
Trained models
After you have finished building your AI model, you're ready to train it. During training, your model learns from your data to perform to your specifications. The time it takes for your model to train depends on the size of your data.
After training is complete, you can view more training and performance information on the details page. In some cases, you can also quick test your model to see it live in action.
Published models
If you're satisfied with your model, you're ready to publish it. You have to publish a model before it can be available to use in Power Apps and Power Automate, so it's important not to miss this step.
Keep track of your model
Any models you create can be accessed at any time by visiting the AI hub page, which you can access from the left-side menu in Power Apps or Power Automate. The AI models page shows a complete list of your models, and their date of latest training and the various states we've learned about here.
Next steps
Now that you have learned how your model can develop over time, your next step is to learn how to edit your model to improve its performance and how to manage multiple model versions.