🎙️ VaaS — Voice as a Service

VaaS is a lightweight, open-source voice processing API built for hackathons and quick prototypes. Get high-quality Speech-to-Text and Text-to-Speech without the headache of cloud billing or complex setups.

🚀 Try the Live Dashboard

🕹️ Features

⚡ 0 Sec Setup: Plug and play with a simple REST API.
🌍 Multilingual: Support for 10+ languages (Spanish, German, Japanese, etc.).
🎙️ Whisper Precise: STT powered by Faster-Whisper.
🔊 Natural Voices: TTS powered by premium Microsoft Edge-TTS.
🛡️ Privacy Focused: All processing is handled on your own Python server.

🤖 AI Agent Integration (Fast-Track)

Paste this into your AI coding assistant (like Antigravity or ChatGPT) to immediately teach it the VaaS API:

The VaaS API is a Voice processing service.
Base URL: https://vaas-production-c42f.up.railway.app
Authentication: Include your key in the 'X-API-KEY' header.

1. Text-To-Speech (TTS):
   - Endpoint: POST /tts
   - Content-Type: multipart/form-data
   - Params: 'text' (string), 'voice' (string, e.g. 'en_us_ava'), 'language' (string, e.g. 'en', 'es').
   - Note: If 'language' is provided without 'voice', a default voice for that language is used.
   - Return: MP3 Audio file.

2. Speech-To-Text (STT):
   - Endpoint: POST /stt
   - Content-Type: multipart/form-data
   - Params: 'file' (audio file/blob), 'language' (string, e.g. 'en', 'hi').
   - Note: 'language' is optional but improves accuracy. Whisper auto-detects by default.
   - Return: JSON { "text": "transcription" }.

🛠️ API Documentation (Detailed)

1. POST `/tts` (Text-to-Speech)

Generates an MP3 file from text.

Parameter	Type	Required	Description
`text`	`string`	Yes	The text to convert to speech.
`voice`	`string`	No	ID of the voice to use (e.g. `en_us_ava`).
`language`	`string`	No	Language code (e.g. `en`, `es`, `hi`) to use a default voice.

Header: X-API-KEY: your_api_key

Example:

curl -X POST "https://vaas-production-c42f.up.railway.app/tts" \
  -H "X-API-KEY: your_key" \
  -F "text=Hello World" \
  -F "language=en" \
  --output voice.mp3

2. POST `/stt` (Speech-to-Text)

Transcribes an audio file into text.

Parameter	Type	Required	Description
`file`	`file`	Yes	The audio file (WAV, MP3, WEBM).
`language`	`string`	No	Explicit language code for transcription (e.g. `en`, `es`).

Header: X-API-KEY: your_api_key

Example:

curl -X POST "https://vaas-production-c42f.up.railway.app/stt" \
  -H "X-API-KEY: your_key" \
  -F "file=@voice.mp3" \
  -F "language=en"

📦 Local Setup

If you want to run VaaS yourself:

# Clone the repo
git clone https://github.com/schallten/vaas
cd vaas

# Setup environment
uv venv
source .venv/bin/activate
uv pip install -r requirements.txt

# Run!
uvicorn vaas.main:app --reload

Built with ❤️ for the developer community. Open source under Apache License 2.0.