OmniVoice · Test Console

▸ Voice Design — Text to Speech

Text to synthesize

Voice Attributes (optional — click tags below or type your own)

male female child teenager young adult middle-aged elderly low pitch moderate pitch high pitch very low pitch very high pitch whisper american accent british accent australian accent canadian accent indian accent chinese accent

Tip: Click pills above to add attributes, or type freely and press Enter/comma to confirm each tag. Language is auto-detected from your text — just write in any language.

Diffusion Steps

Speed

1.00

✓ Audio Generated

↓ Download WAV

▸ Zero-Shot Voice Cloning

Text to synthesize in cloned voice

Reference audio (WAV / MP3 / FLAC · 3–10 seconds)

🎵

Drag & drop or click to upload reference audio

Reference transcript (optional — auto-transcribed by Whisper if blank)

Diffusion Steps

Speed

1.00

✓ Voice Cloned

↓ Download WAV

▸ System Health

API

—

CUDA

—

GPU

—

VRAM Total

—

VRAM Free

—

Model

—

▸ API Reference

Method	Endpoint	Body	Returns
GET	/health	—	JSON — GPU + model status
POST	/synthesize	JSON	audio/wav — voice design TTS
POST	/clone	multipart/form-data + file	audio/wav — voice-cloned speech
GET	/docs	—	Swagger UI (auto-generated)

▸ Voice Attributes Reference

# Gender  male, female
# Age     child, young, middle-aged, elderly
# Pitch   very low, low, normal, high, very high
# Style   whisper
# Accent  american accent, british accent, australian accent

# Example combinations:
instruct = "female, low pitch, british accent"
instruct = "male, elderly, american accent"
instruct = "female, child, high pitch, whisper"

▸ cURL Examples

# Synthesize
curl -X POST http://164.52.194.46:8007/synthesize \
  -H "Content-Type: application/json" \
  -d '{"text":"Hello world","instruct":"female, british accent","num_step":32}' \
  --output output.wav

# Voice clone
curl -X POST http://164.52.194.46:8007/clone \
  -F "text=Hello from OmniVoice" \
  -F "reference_audio=@/path/to/ref.wav" \
  --output cloned.wav

▸ Python Example

import requests

# Synthesize with voice attributes
r = requests.post(
  "http://164.52.194.46:8007/synthesize",
  json={"text": "Hello", "instruct": "female, british accent", "num_step": 32})
open("out.wav", "wb").write(r.content)

# Voice clone
r = requests.post(
  "http://164.52.194.46:8007/clone",
  data={"text": "Hi there"},
  files={"reference_audio": open("ref.wav", "rb")})
open("cloned.wav", "wb").write(r.content)