ElevenLabs
Very natural AI voices: narrate texts, clone voices, translate videos and transcribe speech.
What it does
ElevenLabs reads texts aloud with voices that capture emphasis and emotion remarkably well – in around ninety languages, including German with several accents. A few minutes of recording are enough to clone a particular voice. A dubbing studio translates videos while keeping the speakers’ voices. On top come speech recognition, sound effects, music and an API for building voice assistants for phone lines or websites.
Who it suits
Audio versions of articles, explainer videos, e-learning, podcasts, product videos in several languages, and development teams adding voice features to their own applications.
Data protection for businesses
In the standard plans, inputs are used to improve the models; you can switch this off in your profile under data use. EU data storage and a zero-retention mode are available only on the Enterprise plan. Other people’s voices may only be cloned with their consent.
Limits
The free tier does not allow commercial use. Billing uses credits that are consumed at different rates depending on the model. Technical terms and names sometimes need pronunciation hints.