Transcribeer

The modern, multilingual audio-to-text transcription tool powered by Vosk.
Cross-platform for Windows and Linux, with beautiful UI and AI-powered (GEMINI API REQUIRED) text enhancement.

Features • Download and Install • How It Works • Languages & Models • AI Enhancement • License

---

🚀 Features

🎧 Audio Transcription

• Multi-format Support: MP3, WAV, M4A, OGG, FLAC
• Automatic Conversion: Converts audio to optimal format for speech recognition
• Real-time Progress: Live progress bar and status updates

🌍 Multilingual Support

• 8 Languages: Italian, English, French, German, Spanish, Portuguese, Russian, Chinese
• Auto-language Detection: UI automatically matches your system language
• Model Auto-download: Downloads speech recognition models on first use

🤖 AI-Powered Enhancement

• Google Gemini Integration: Improves punctuation and text flow
• Smart Punctuation: Automatic sentence capitalization and punctuation
• Text Refinement: Enhances readability while preserving meaning

🎨 Modern Interface

• Dark Theme: Easy on the eyes modern UI
• CustomTkinter: Beautiful, customizable interface
• Intuitive Workflow: Simple and user-friendly

⚡ Technical Features

• Offline Capable: Works without internet (except AI features)
• Cross-Platform: Windows and Linux support
• Lightweight: Fast and efficient processing

---

📦 Download & Installation

Windows

Requirements:

• Windows 10 or later
• IMPORTANT NOTE: FFmpeg (winget install Gyan.FFmpeg)

Installation:

1. Install the latest version Transcibeer: 👉 Latest Windows Release
2. Download FFmpeg by opening the command prompt/powershell and entering the following line of code: (winget install Gyan.FFmpeg).
3. When downloading FFmpeg, Press Y to confirm the download and wait for it to be completed.
4. Run Transcibeer with administrator.
5. In Transcibeer, select a language model and download it.

Linux

Requirements:

• Ubuntu/Debian: sudo apt install ffmpeg
• Fedora: sudo dnf install ffmpeg
• Arch: sudo pacman -S ffmpeg Transcribeer: 👉 [Latest Linux Release] - No Releases for now. Coming soon.

From source code folder (dev only)

---

Extra information

Clone repository

git clone https://github.com/il-mangia/
cd Transcribeer

Install dependencies

pip install -r requirements.txt

Run application

python main.py

---

🧠 How It Works

1. FFmpeg converts your audio to WAV 16 kHz mono
2. Vosk (https://alphacephei.com/vosk/) transcribes and auto-detects the language
3. Transcript is translated into italian only
4. Both original and translated text are shown
5. You can save everything into a .txt file

---

🌍 Languages & UI

Supported languages:

• 🇮🇹 Italian (it) - Vosk Small IT 0.22
  
• 🇺🇸 English (en) - Vosk Small EN-US 0.15
  
• 🇫🇷 Francais (fr) - Vosk Small FR 0.22
  
• 🇩🇪 Detusch (de) - Vosk Small DE-Zamia 0.3
  
• 🇪🇸 Spanish (es) - Vosk Small ES 0.42
  
• 🇵🇹 Portoughese (pt) - Vosk Small PT 0.3
  
• 🇷🇺 Russian (ru) - Vosk Small RU 0.22
  
• 🇨🇳 Chinese (cn) - Vosk Small CN 0.22
  

---

🤖AI Enhancement

Transcribeer can enhance your transcripts using Google's Gemini AI, to implement this feature follow this procedure:

Get API Key:

1. Visit the Google AI Studio.
2. Log in your Google Account and generate an API key.

Configuration in App:

4. Click "AI" button in main interface
5. Enter your API key when prompted
6. They Key will be saved locally for future use

What AI Improves:

1. Punctuation: Adds proper periods, commas, question marks
2. Capitalization: Correct sentence case
3. Readability: Improves text flow and naturalness
4. Preservation: Maintains original meaning and content

🔧 Development Setup (Source code folder)

Install dependencies:

pip install -r requirements.txt  

Run:

python main.py  

---

🧪 Supported media Formats

• MP3
• WAV
• AAC
• M4A

(All are converted to WAV automatically.)

🧰 Tech Stack

• Python 3
• Vosk
• Customtkinter
• FFmpeg
• Google Gemini API

📝 Known Limitations (Important)

• AI function requires Internet.
• GPU acceleration planned for future versions.
• Local translation model planned.

❤️ License

OPEN SOURCE!!!!!!!!!

---

Built with ❤️ by Il-Mangia — Powered by Whisper