KeepYourMouthShut - Acid Reflux for your Ears!

       

[Please consider sponsoring this project to help keep it alive!]

KeepYourMouthShut is an AI-assisted open-source app that provides a 360-degree podcast solution 🎙️ It has a developer-first approach for creating entire episodes of podcasts without requiring any expensive equpment or setup. The app takes a number of input from the end-user and based on that generates a script and an audio voice-over file based on that script.

Table of Content

• 🚀 Installation
• ⚠️ Deployment
• 🛠️ Built With
• 🎸 Music
• 🚸 Contributing
• 🪫 Versioning
• 🏅 Acknowledgement
• 🕹️ License
• 📟 Citation

🚀 Installation

For local installation, clone the repository and run the following commands:

1. Install FFMPEG
2. Create a free account on llmOS and get an API key
3. Insert the API key in the .env file. A sample .env file is given.
4. Install uv and Node.js
5. Run uv sync
6. Run uv run uvicorn main:app --app-dir app --host 127.0.0.1 --port 8000
7. In another terminal, from frontend/, run npm install and then npm run dev

The Next.js app is at http://127.0.0.1:3000. It sends the form to the FastAPI backend on port 8000 and downloads the generated zip. Set NEXT_PUBLIC_API_URL if the API is not at http://127.0.0.1:8000, and set CORS_ORIGINS on the API to the UI origin.

⚠️ Deployment

Serve the API with uvicorn and the UI with npm run build followed by npm run start inside frontend/. Point NEXT_PUBLIC_API_URL at the public API origin, and set CORS_ORIGINS to the public UI origin.

🛠️ Built With

• Python
• FastAPI — API
• Next.js — web UI
• uv — Python package manager
• llmOS — default LLM (Mistral-7B)
• edge-tts — default TTS (en-GB-SoniaNeural)
• 60db — optional drop-in peer for TTS + LLM + STT (this fork)
• Pixabay — royalty-free background music

60db provider (alongside edge_tts + llmOS)

This fork adds 60db as a peer provider for the same two services KYMS already uses. One SIXTYDB_API_KEY covers all three concerns; each switch is independent so you can mix providers:

Concern	File	Env switch	60db endpoint
TTS	app/utils/sixtydb_tts_stuff.py	TTS_PROVIDER=sixtydb	POST /tts-synthesize (default), POST /tts-stream, or wss://api.60db.ai/ws/tts
LLM	app/utils/sixtydb_llm_stuff.py	LLM_PROVIDER=sixtydb	POST /v1/chat/completions
STT	app/utils/sixtydb_stt_stuff.py	utility (no auto-consumer in pipeline)	POST /stt

Default behaviour is unchanged — both edge_tts and llmOS keep working without any .env edits. Flip the switches when you want 60db.

Pick the TTS transport surface via SIXTYDB_TTS_TRANSPORT=sync|stream|ws (default sync; ws requires the optional websockets package).

🎸 Music

• Whistle Vibes - Music by Sergio Prosvirini from Pixabay
• Scandinavianz Thessaloniki - Music by Oliver Moilanen from Pixabay
• Lofi Chill (medium version) - Music by Bohdan Kuzmin from Pixabay
• Bolero - Music by Oleksii Kaplunskyi from Pixabay

🚸 Contributing

Please read CONTRIBUTING.md for details on our code of conduct, and the process for submitting pull requests to us.

🪫 Versioning

We use SemVer for versioning. For the versions available, see the tags on this repository.

🏅 Acknowledgement

1. Adam Tal - https://github.com/AdmTal
2. rany2 - https://github.com/rany2

🕹️ License

This project is licensed under the GPL-3.0 license, see the LICENSE.md file for details

Star History

📟 Citation

cff-version: 1.2.0
message: "If you use this software, please cite it as below."
authors:
- family-names: "Bhattacharjee"
  given-names: "Rajtilak"
  orcid: "https://orcid.org/0009-0008-2465-0266"
title: "keepyourmouthshut"
version: 0.0.1
doi: 10.5281/zenodo.10199470
date-released: 2023-11-23
url: "https://github.com/rajtilakjee/keepyourmouthshut"