Select a voice to use it in Generate speech.
Your cloned voices
Clone a voice
Default voices
No voice selected
0 / 300
Higher = more expressive
Lower = slower pacing
Routes via CF Worker → HF Space
Reference audio
Drop WAV or MP3 here
10–20 seconds · clean speech · no music
Tips for best quality
• Speak naturally — no reading hesitations
• Silence before & after the clip helps
• 44.1 kHz WAV gives cleanest result
• Match language to what you'll generate
• Silence before & after the clip helps
• 44.1 kHz WAV gives cleanest result
• Match language to what you'll generate
Voice profile
English
Urdu
Narration
Conversational
Formal
Fast
Slow
No generations yet — generate some speech first.
CF Worker endpoint
Default generation settings