Podcastfy: An Open Source Python Tool for Turning Content into Podcasts
Ever wished you could turn that long article, research paper, or even a YouTube transcript into a podcast? Podcastfy is here to help—an open-source Python alternative to NotebookLM’s podcast feature that transforms text (or even multimodal content) into engaging multilingual audio conversations using generative AI.
No proprietary APIs, no paywalls—just a Python package that lets you automate podcast creation from any text source. Whether you're a content creator, developer, or just someone who prefers listening over reading, Podcastfy makes it easy to repurpose content for audio consumption.
What It Does
Podcastfy takes text input (or extracts it from PDFs, web pages, or videos) and converts it into a natural-sounding podcast-style conversation. It uses generative AI to structure the content into a dialogue format, then leverages text-to-speech (TTS) models to produce high-quality audio output.
Key features:
- Multilingual support – Generate podcasts in multiple languages.
- Multispeaker TTS – Simulate a conversation with different AI voices.
- Docker & API support – Deploy it locally or integrate it into your workflow.
- Open source & extensible – Swap TTS models, tweak prompts, or contribute new features.
Why It’s Cool
Unlike most podcast-generation tools, Podcastfy is entirely self-hostable and doesn’t lock you into a specific AI provider. It supports Google’s TTS models, OpenAI, and more, so you can pick the best voices for your needs.
The real magic? It doesn’t just read text monotonously—it restructures content into a conversational format, making it feel like a real podcast discussion. Imagine turning a dense whitepaper into an engaging audio explainer without manual editing.
How to Try It
- Install via pip:
pip install podcastfy - Run it on a text file:
podcastfy generate --input my_article.txt --output podcast.mp3 - Or use Docker for a full setup:
docker-compose up
Check out the GitHub repo for full docs, or try the web demo if you want a quick test drive.
Final Thoughts
Podcastfy is a great example of how generative AI can automate content repurposing in a useful way. Developers can integrate it into workflows (think automated podcast feeds from blog posts), while content creators can use it to expand their reach into audio formats.
It’s not perfect—TTS still has quirks, and the conversational flow depends on the source material—but as an open-source project, it’s a solid starting point for anyone who wants to experiment with AI-generated audio.
Got ideas for improving it? The project’s actively maintained, so feel free to contribute!