Start small. Listen more when you need to.

Create and edit scripts without signing up, then preview up to 60 seconds. Guests receive one audio render per episode; signed-in accounts receive 3 renders per month, up to 10 minutes each. Regeneration counts. Daily and site-wide limits also apply.

PDF, DOCX, PPTX and UTF-8 TXT, up to 10 MB, 100 PDF pages and 60,000 extracted characters. Scanned PDFs need OCR first. English and Chinese are supported. Paste webpage text or YouTube captions into the text tab. Automatic URL import is being prepared.

This app does not train models on your documents. Files are parsed in your browser. Extracted text, scripts and audio are retained for up to seven days, then refused and cleaned on subsequent requests; audio expires automatically. Text is sent to the selected AI provider under its data terms. You can delete an episode sooner.

You must have rights to the source material and comply with the selected AI and voice provider terms. The included sample recordings are for product evaluation. Check provider terms before commercial distribution.

AI can make mistakes. Inspect the linked source excerpts before voicing. Without a model key, the app creates an explicitly labeled extractive outline instead of AI paraphrases. Edited lines should be checked again.

Create a private HTTPS RSS link while signed in, then add it to apps such as Apple Podcasts or Overcast. Spotify does not accept private RSS. Anyone with your feed link can listen; reset it to revoke access. Episodes expire after seven days.