OmniVoice Studio
Diffusion-based zero-shot voice cloning & synthetic voice design supporting 600+ languages powered by OmniVoice Discrete Diffusion nodes in ComfyUI.
⚡ 20s GPU Diffusion
🌐 600+ Languages
🧬 Pure Voice Design (No Audio Required)
🎭 Dialect & Accent Control
1
Reference Voice Sample
Provide 3 to 15 seconds of clean, single-speaker audio
Drag & Drop Reference Audio
WAV, FLAC, MP3, OGG, or WEBM (up to 20MB)
Auto-transcribed with Whisper if blank
Quick Style Presets:
British Female Narrator
Deep American Male
Energetic Female Podcaster
Elderly Storyteller
Soft Whisper Voice
Australian Male
Canadian News Anchor
Movie Trailer Voice
2
Target Speech Script
Text to synthesize in the cloned or designed voice
Supports 600+ languages natively
0 characters
3
Diffusion Parameters
Fine-tune model precision, sampling steps, guidance, and pacing
⚡ Discrete diffusion on RTX 3060 CUDA
Leave blank to inherit from reference audio
16 (Fast ~12s)
32 (Balanced ~20s)
64 (Studio)
1.0 (Creative)
2.0 (Standard)
4.0+ (Strict)
Estimated time: ~15 to 25 seconds on RTX 3060 CUDA GPU
✓
Synthesized Audio Master
Diffusion-generated vocal master with high-fidelity acoustic reproduction
No Voice Generated Yet
Select your mode (Voice Clone or Voice Design), enter your script, and press Synthesize with OmniVoice Diffusion to start.
💡 Tip: You don't need reference audio! Switch to Synthetic Voice Design to invent voices purely from text descriptions (gender, age, pitch, accent).
Recent Generations
Generated audio masters from this session will appear here.
Server Storage Management: