MCP Audio Transcriber

A Dockerized Python tool that implements the Model Context Protocol (MCP) via AssemblyAI's API. Upload or point to an audio file, and receive a structured JSON transcription.

Features

• AssemblyMCP: a concrete MCP implementation that uses AssemblyAI's REST API

• Command-line interface (app.py):

  python app.py <input_audio> <output_json>

• Streamlit web UI (streamlit_app.py):

  • Upload local files or paste URLs

  • Click Transcribe

  • Preview transcript and download JSON

• Docker support for environment consistency and portability

Related MCP server: Transcripter MCP Server

Prerequisites

• Python 3.10+

• An AssemblyAI API key

• ffmpeg (for local decoding, if using local files)

• (Optional) Docker Desktop / Engine

• (Optional) Streamlit (pip install streamlit)

🔧 Installation

1. Clone the repo

   git clone https://github.com/ShreyasTembhare/MCP---Audio-Transcriber.git cd MCP---Audio-Transcriber

2. Create a .env

   ASSEMBLYAI_API_KEY=your_assemblyai_api_key_here

3. Ensure .gitignore contains:

   .env

4. Install Python dependencies

   pip install --upgrade pip pip install -r requirements.txt

5. Install ffmpeg

   • Ubuntu/Debian: sudo apt update && sudo apt install ffmpeg -y

   • Windows: download from https://ffmpeg.org and add its bin/ to your PATH

Usage

1. CLI Transcription

• <input_audio>: any file or URL supported by AssemblyAI

• <output_json>: path for the generated JSON

Example:

2. Streamlit Web UI

• Open http://localhost:8501

• Upload or enter an audio URL

• Click Transcribe

• Download the JSON result

3. Docker

Build the image:

Run it (mounting your data/ folder):

Then inspect:

Windows PowerShell:

Project Structure