My summary of Eleven Labs before using it — ElevenLabs turns text into voice.
How does ElevenLabs explain itself in the first minute? On its website ElevenLabs explains that its users can create natural AI voices instantly in any language.
How does ElevenLabs work? I select a language — English — and sample text is generated. I then use a dropdown of names to generate a voice. There are additional features such as being able to clone your own voice and ‘Speech to Speech’.
Once you create an ElevenLabs account, you get access to a Speech Synthesis page where you can select a range of voices. The settings of each voice can be customised. For example, you can increase the stability of the voice to make it more monotone or decrease stability to add more expression. Or when you boost the clarity and similarity of the voice, you’ll make the voice sound more like the original.
ElevenLabs has its own Voice Library from which users can add specific voices to their own ‘Voice Lab’ and further customise these voices.
What are the use cases ElevenLabs aims to solve for? ElevenLabs is focused on content creation use cases. Short story writers, for example, can use ElevenLabs to convert their written text into audio content. Or game creators can use ElevenLabs to add narration or dialogue to the games.
ElevenLabs pricing plans are based on elements such as the number of characters per month, audio quality and license to use ElevenLabs for commercial use.
Main learning point: I really like the interface and the range of functionality that ElevenLabs offers for AI voice generation. Some of the voices didn’t sound particularly realistic to me, and I can imagine that ElevenLabs will offer more functionality in future to further tune the expression, pitch and diction of each voice.
Related links for further learning:
