This guide will help you get started with ElevenLabs. We will cover everything, starting with creating an account, and then moving to cloning your first voice and generating your initial voiceover. We will also cover prompting techniques (how to influence the AI’s performance) as well as its current limitations and challenges.

We will guide you through the various stages of ElevenLabs, starting with VoiceLab; this is where you can create or clone voices according to your preferences. Once you have set up your desired voices, we will move on to Speech Synthesis. Here, you will be able to generate your first audio outputs using the pre-made voices or the ones you’ve created or cloned.

How does the AI model work?

The AI has been trained on a vast amount of audio. The type of audio varies, but the most prominent is audiobooks. This is the context it understands the best, and it provides the most predictable results when generating audio. If you write something in the style of a book, the AI can sometimes interpret how to perform a certain passage from the context of the writing itself. To achieve a more emotive range, you can lower the stability slider, although this may sacrifice some degree of predictability.

With each successive update to the model, where it has been re-trained, the AI gets better and better at understanding different contexts as its dataset grows. This will help it understand more nuances between humans, languages, and accents.