ElevenLabs Unveils GenFM: Transforming Text and Videos into Engaging AI-Generated Podcasts

On Wednesday, ElevenLabs enhanced its ElevenReader app by introducing a feature that leverages artificial intelligence to create instant audio content similar to podcasts. Known as GenFM, this function enables users to input any text or link to a YouTube video, transforming it into an engaging audio experience with two AI-generated hosts sharing insights and intriguing information drawn from the original content. Presently, this feature can be accessed at no cost on the iOS version of the app, supporting audio generation in 32 different languages.

In contrast, Google’s NotebookLM, launched in June, allows users to create AI-generated overviews of various written materials in a podcast style, featuring two AI hosts discussing the content. Nevertheless, it is solely available as a web application and currently supports only English, presenting a limitation for some users. The GenFM feature from ElevenLabs effectively addresses these limitations by integrating multilingual support into its AI podcast functionality, making it available in languages such as Hindi, Portuguese, Chinese, Spanish, French, German, Japanese, and many others.

At this time, the feature is exclusive to the iOS app, but plans are in motion to make it available on the Android platform in the near future. To employ this new feature, users can easily paste text, upload documents, or provide the URL of a YouTube video, allowing the platform to seamlessly convert the input into a conversational podcast format. The app intelligently selects two voices from an array of over 12 options, enabling the AI hosts to read the content along with noteworthy insights in a manner that mimics human interaction.

The technology behind this feature relies on ElevenLabs' advanced AI audio models, which facilitate podcast creation in mere seconds. However, specific details regarding the AI models and the sources of training data have not been disclosed. Importantly, the company reassures users that it does not retain any personal data after the podcast is generated. 

The ElevenReader app stands out as an AI-driven text-to-speech tool capable of handling various formats, including PDF, ePUB, URLs, and even text found within images. Available for free, users also have the option to publish their audio outputs for others in the community to enjoy, although the audio quality leans towards a more synthetic sound.