Spaces:
Sleeping
Sleeping
| title: "🎙️ Voice Cloning Studio" | |
| emoji: 🎙️ | |
| colorFrom: purple | |
| colorTo: blue | |
| sdk: gradio | |
| sdk_version: "5.39.0" | |
| app_file: app.py | |
| pinned: false | |
| short_description: "Free voice cloning with F5-TTS + Zonos emotions" | |
| tags: | |
| - voice-cloning | |
| - tts | |
| - text-to-speech | |
| - f5-tts | |
| - zonos | |
| - emotional-tts | |
| # 🎙️ Voice Cloning Studio | |
| **Clone any voice for free** using the world's best open-source TTS models. | |
| ## Features | |
| | Tab | Model | What it does | | |
| |-----|-------|-------------| | |
| | 🎙️ **Voice Clone** | F5-TTS | #1 voice cloning — upload 5-15s audio, get any text in that voice | | |
| | 🎭 **Emotional Voice** | Zonos | 8 emotion sliders (happy/sad/angry/scared/surprised/disgusted/neutral) + voice cloning | | |
| | ⚡ **Compare** | Both | Side-by-side comparison of both models | | |
| | 📖 **Guide** | — | Complete TTS model landscape & tips | | |
| ## How it works | |
| This app is a **free wrapper** around existing HuggingFace Spaces: | |
| - [F5-TTS](https://huggingface.co/spaces/mrfakename/E2-F5-TTS) — best zero-shot voice cloning | |
| - [Zonos](https://huggingface.co/spaces/Steveeeeeeen/Zonos) — best emotional control in TTS | |
| **No GPU needed** — all heavy computation runs on the backend Spaces for free. | |
| ## Models used | |
| - **F5-TTS** ([paper](https://arxiv.org/abs/2410.06885)) — DiT + Flow Matching, 335M params, MIT license | |
| - **Zonos v0.1** ([Zyphra](https://zyphra.com)) — DAC codec + Transformer, 44kHz output, Apache 2.0 | |