README.md

---
title: Audiotranscribe
emoji: 📝
colorFrom: red
colorTo: purple
sdk: gradio
sdk_version: 6.12.0
app_file: app.py
pinned: false
license: apache-2.0
short_description: Transcribe Audio and Create a Summary with Action Items
---

# 🎙️ Audio Transcription & Summary App

A lightweight, streamlined AI web application that automatically transcribes audio files or microphone recordings, and instantly generates a concise summary along with extracted action items.

Try it live on Hugging Face Spaces: [**Audiotranscribe**](https://huggingface.co/spaces/madanyc/audiotranscribe)

## ✨ Features

- **Flexible Input:** Upload existing audio files (`.mp3`, `.wav`, `.m4a`, etc.) or record directly from your microphone in the browser.
- **Fast Transcription:** Uses Hugging Face Inference API with the `openai/whisper-large-v3-turbo` model for high-accuracy speech-to-text.
- **Smart Summarization & Action Items:** Passes the transcript to `meta-llama/Llama-3.3-70B-Instruct` to generate a quick summary and automatically pull out action items/deadlines in a single, optimized LLM call.
- **Downloadable Output:** Export the transcript, summary, and action items together as a `.txt` file for easy sharing or archiving.
- **Clean UI:** Built with Gradio for a simple, single-page user experience.

## 🛠️ Requirements & Tech Stack

- Python 3.10+
- [Gradio](https://www.gradio.app/) (UI Framework)
- [Hugging Face Hub](https://huggingface.co/docs/huggingface_hub/index) (API Client)
- `python-dotenv`

## 🚀 Running Locally

1. **Clone the repository** (or download the files).
2. **Install dependencies:**
   ```bash
   pip install -r requirements.txt
   ```
3. **Set up your environment variables:**
   Create a `.env` file in the root directory and add your Hugging Face token. You can get one from [hf.co/settings/tokens](https://huggingface.co/settings/tokens). It must have read access for the inference API.
   ```
   HF_TOKEN=your_hugging_face_token_here
   ```
4. **Run the app:**
   ```bash
   python app.py
   ```
5. **Open your browser:** The app will be available at `http://127.0.0.1:7860` (or similar, depending on your setup).

## ☁️ Deployment

This app is configured to be easily deployed to [Hugging Face Spaces](https://huggingface.co/spaces). 
The `app.py` script and the YAML block at the top of this `README.md` are all that's needed. Ensure you add `HF_TOKEN` to your Space's Secrets in the Settings tab.