NotebookLM facilitates the creation of podcasts featuring two people based on the prompt and data-source we provide. However, a significant limitation is that it currently supports only English. I’m keen to explore ways or platforms (even if it requires a multi-platform approach) to generate podcasts between two AI personalities in other languages (like German, Hindi, Arabic) as well.
In the past, I’ve used ElevenLabs to generate audio from scripts. However, its capabilities are limited to scripts with a single speaker. Are there platforms that can handle scripts with multiple speakers?
For example, let’s say I have a script generated by ChatGPT with two speakers taking turns in dialogue. I want to process this script through TTS platforms to produce meaningful podcasts. Or maybe I want to be able to give the platform the some sources to generate podcasts (similar to NotebookLM does) but in non-English languages. My primary requirement is that the speakers should have distinct voices to differentiate them. Is this possible?
Are there any ideas or solutions to achieve this? Any recommendations for platforms or workflows that could help with such a setup would be greatly appreciated.