Hosts and guests detected
Works with common formats like "Host: line" or "[Host]: line". If your script isn't labeled, AI can detect the speakers for you.
AI Podcast Generator
Turn your script into a two-host podcast
Paste or import your script and we detect the host and the guest. Give each one your cloned voice or a voice from the library, generate section by section, redo only the lines you don't like, then export the full episode as MP3.
No studio, no editing software to learn
One voice per speaker
Pick your cloned voice or a public library voice for each speaker. Changing a voice only affects that speaker's sections.
Generate by section, redo one line
Preview, regenerate and keep earlier versions of every section. Fixing one line never means regenerating the whole episode.
Control pauses and pace
Write [pause 800ms] in your script for breaths and transitions, and adjust speed or mode for any single section.
Listen to the whole episode
Play generated sections back to back. When something sounds off, jump straight to that section.
Export files ready to publish
Download the full episode MP3, a ZIP of section MP3s and the script, ready for your podcast host or editor.
Interviews and explainers
Turn an article or interview notes into a conversation that's easier to listen to.
Brand and product podcasts
Keep publishing with the same host voices so your show always sounds like you.
Courses and training
Turn handouts into Q&A style audio lessons people can listen to on the go.
Multilingual editions
Swap in a translated script and make another edition with the same voices.
- 1
Prepare the script
Paste it, or import Word, TXT, spreadsheet or subtitle files. You can also start from a solo or two-host example.
- 2
Assign voices
Choose a cloned or public voice for each host. You'll see the credits needed before generating.
- 3
Generate and export
Generate everything, review section by section, redo only what you don't like, then export the full MP3.
No. You can use public voices from the Voice Square right away. To use your own voice, record a 10–30 second sample once and reuse it in every episode.
Make your first episode now
Start from an example script, replace it with your own and hear the result in minutes.