Alibaba's Voice Revolution: Speak It Into Existence

Imagine telling your computer "Make this voice sound like a nervous teenager" or "Add café chatter in the background" - and having it happen instantly. That's the promise of Alibaba Tongyi Lab's new voice technology duo unveiled today.

Your Personal Voice Director

The Fun-CosyVoice3.5 isn't your average text-to-speech tool. Want your audiobook narrator to sound more dramatic? Just say "Add some Shakespearean flair." Need customer service training audio? Tell it to "sound patient but slightly exasperated." This multilingual whiz now understands Thai, Indonesian, Portuguese and Vietnamese too - with obscure character errors slashed by nearly 70%.

Meanwhile, Fun-AudioGen-VD acts like a Hollywood sound studio in your browser. Picture this:

• "Create a deep-voiced villain with a slight lisp standing in a cathedral"
• "Make a children's storyteller with background forest sounds"
• "Simulate an underwater conversation between two robots"

The system handles everything from subtle vocal quirks to complex environmental acoustics.

Why This Changes Everything

For podcasters, these tools eliminate expensive voice actors for placeholder tracks. Game developers can prototype character voices before recording sessions. Even filmmakers can quickly generate temporary dialogue during editing.

"We're removing the technical barriers," explains Tongyi Lab's spokesperson. "Now creative vision directly translates to audio reality."

The models aren't perfect yet - extremely specific requests might still need tweaking. But for most users, speaking their audio needs into existence just became reality.

Key Points:

• Natural language control: Adjust voices and scenes using everyday phrases
• Multilingual mastery: Supports 13 languages with improved accuracy
• Lightning fast: 35% reduction in processing delays
• Creative playground: Combine characters, emotions and environments freely