Podify: A Small, Useful AI Podcast Studio
Not a giant platform. Not a complicated production suite. Just a focused tool that helps users turn an idea into something they can listen to.
Podify is a local AI podcast studio. You give it a topic, it researches the idea, writes a podcast-style script, lets you choose voices, and generates a finished audio episode from one Gradio app.
Type an idea. Review the script. Pick voices. Generate the podcast.
That is Podify.
The Idea
Podcasts are one of the easiest ways to learn, explore, and share ideas. But creating one usually takes a lot of work: research, writing, recording, editing, and mixing.
Podify makes that workflow smaller.
Instead of building a large, complex content tool, we focused on one useful user journey:
Topic -> Research -> Script -> Voices -> Podcast
The result is a compact app that feels practical. It is small enough to run locally, but complete enough to produce something real.
Local-First and User-Focused
The focus of Podify is not just automation. It is usefulness.
A user should be able to sit down with an idea and quickly turn it into a listenable podcast draft. That makes Podify useful for learners, creators, students, researchers, and anyone who wants to explore a topic in audio form.
Podify is built as a self-contained Gradio app, so it can run on your own machine. The research and scripting flow can run locally, and the audio generation can run on a local GPU or on Hugging Face Spaces with ZeroGPU.
That local-first direction matters to us. Small AI apps should not always need to become big cloud products. Sometimes the best experience is a focused tool that runs close to the user.
How It Works
Podify has two main phases.
First, it researches and writes. A LangGraph workflow plans the topic, searches the web, gathers context, builds an outline, and writes a speaker-tagged podcast script.
Then, it produces audio. The script is converted into speech using Fish Audio / OpenAudio S1-mini, with support for preset voices and voice cloning from an uploaded clip or microphone recording.
Everything is wrapped in a polished Gradio interface designed to feel like a small podcast studio.
Why Podify Fits Build Small
For us, “build small” means building something focused enough to understand immediately, but useful enough that people can actually use it.
Podify does one thing clearly: it turns a topic into a podcast.
Inside that simple idea, it connects research, agents, scripting, voice selection, speech synthesis, and audio stitching into one smooth workflow.
Small app. Useful output. Runs close to the user.
Features
Podify includes:
- Topic-to-podcast generation
- Research-powered script writing
- Multiple episode styles like interview, storytelling, debate, news, and educational
- Editable speaker-tagged scripts
- Preset voice library
- Voice cloning from audio upload or microphone recording
- Multi-speaker podcast generation
- Background music beds
- Finished podcast audio output
- A dark studio-style Gradio UI
- Local-first app structure that can run on your own device
Built With
Podify uses:
- Gradio for the app interface
- LangGraph for the research and writing flow
- Hugging Face Inference for the content model
- DuckDuckGo search for live research
- Fish Audio / OpenAudio S1-mini for text-to-speech
Try Podify
Podify is our take on what a small AI app can be: focused, useful, creative, and close to the user.
It helps you go from idea to audio in minutes.
Demo: https://huggingface.co/spaces/build-small-hackathon/podify
#build-small-hackathon #backyard-ai #gradio #text-to-speech #podcasts
