---
title: Mazinger Dubber
emoji: "\U0001F3AC"
colorFrom: yellow
colorTo: green
sdk: gradio
sdk_version: 5.50.0
app_file: app.py
pinned: true
license: mit
preload_from_hub:
  - openai/whisper-large-v3
tags:
  - video-dubbing
  - voice-cloning
  - translation
  - qwen3-tts
  - whisperx
  - arabic
---

# Mazinger Dubber

End-to-end video dubbing studio powered by [Mazinger](https://github.com/bakrianoo/mazinger).

**Pipeline:** Download → Transcribe (WhisperX) → Thumbnails → Describe → Translate → Resegment → Synthesize (Qwen3-TTS) → Assemble → Subtitle

**Features:**
- Voice themes, pre-built profiles, or custom voice cloning
- Multi-language translation via HF Inference API
- Burned subtitles with Google Fonts support
- No API keys needed — fully powered by HF infrastructure

**Limitations:**
- Max ~5 min video audio (ZeroGPU constraint)
- First run may be slower due to model downloads

Built by [@HeshamHaroon](https://huggingface.co/HeshamHaroon) | Powered by [Mazinger](https://github.com/bakrianoo/mazinger) by [@bakrianoo](https://github.com/bakrianoo)