ElevenLabs v4 Changes the Voice AI Game, and I’m Proud I Chose AI Early

By Pocholo De Leon Gonzales, The VoiceMaster of the Philippines

The future of the human voice is changing again.

On September 28, 2026, ElevenLabs introduced ElevenLabs v4 and v4 Turbo, its newest generation of speech models. The update brings stronger emotional expression, greater control over delivery, faster voice cloning, lower latency for conversational AI, and support for more than 90 languages.

For most people, this is another major AI announcement.

For me, it is something much more personal.

I have spent almost three decades working with the human voice. I grew up in radio. I became a voice actor, dubbing director, trainer, producer, speaker, and eventually an advocate for Filipino voices in artificial intelligence.

And while many people in the traditional voice industry were afraid of voice cloning, I made a different decision.

I accepted AI.

I studied it.

I experimented with it.

I allowed my voice to become part of it.

And I made it my mission to make sure that Filipino voices would not be left behind.

ElevenLabs v4 Is Not Just About Sounding Human

The biggest change in v4 is not simply better audio quality.

It is control.

AI speech is moving from merely reading words correctly to understanding how those words are supposed to feel.

With v4, ElevenLabs says the system can maintain voice identity across longer passages while considering the context of the text to determine how a line should be delivered. Expression controls introduced in v3 are also being expanded, allowing multiple inline expression tags to be combined and followed sequentially.

That means creators can increasingly direct AI voices almost like human performers.

Whisper.

Pause.

Laugh.

Become nervous.

Become excited.

Change emotional intensity.

Shift delivery as the scene develops.

For someone who has spent decades directing voice actors, this development is significant.

We are entering an era where prompting becomes part of voice direction.

The voice director of tomorrow will not only tell an actor what emotion to perform. That director may also be directing an AI model through language, context, timing, emotional instructions, and performance tags.

That changes the craft.

Voice Cloning From Just Seconds of Audio

Another striking development is speed.

ElevenLabs says Instant Voice Cloning can work with as little as 10 seconds of audio, while its documentation confirms that short samples can be used to create an Instant Voice Clone.

Think about what that means.

Voice cloning used to sound like something requiring laboratories, engineers, huge datasets, expensive equipment, and weeks of training.

Now we are talking about seconds.

That dramatically lowers the barrier to creating personalized synthetic voices.

It also makes one thing even more important:

consent.

The technology is becoming easier. Therefore, responsible voice ownership, licensing, authorization, attribution, and protection must become stronger.

I am not against voice cloning.

I am against unauthorized voice cloning.

There is a huge difference.

Technology itself is not the enemy. Abuse is.

More Than 90 Languages Means the AI Voice Revolution Is Becoming Global

Perhaps the development closest to my heart is ElevenLabs' expansion beyond 90 languages.

Filipino is already supported across several ElevenLabs voice technologies, including voice cloning, and the company's broader dubbing platform now supports more than 90 languages.

This matters because artificial intelligence cannot truly become global if everyone sounds American.

The world does not speak with one accent.

We have Filipino.

Cebuano.

Hiligaynon.

Waray.

Kapampangan.

Ilocano.

Bicolano.

Chavacano.

And hundreds of other languages, dialects, accents, speech patterns, cultural expressions, and ways of telling stories.

Voice AI should not erase those identities.

It should preserve them.

That is why I entered this industry.

I Chose to Become Part of AI Instead of Fighting It

When generative AI and voice cloning began accelerating, many voice professionals saw only one possibility:

AI versus voice actors.

I saw another possibility:

AI with voice actors.

I believed that if voice technology was going to learn how humanity speaks, then Filipinos needed to be inside that development.

Not watching from the sidelines.

Not complaining after the technology was finished.

Not asking years later why Filipino voices sounded unnatural.

We needed to participate.

So I did.

I worked with emerging AI voice technologies. I experimented with cloning my own voice. I collaborated with companies building speech and synthetic voice systems. I contributed Filipino speech, pronunciation, performance knowledge, cultural context, and professional voice experience.

I openly embraced something many people in my own industry wanted nothing to do with.

And I make this claim proudly:

I became the first Filipino professional voice artist to openly embrace AI voice cloning as part of the future of the voice industry, instead of campaigning against it.

I did not accept AI because I wanted to replace Filipino voice artists.

I accepted it because I wanted Filipino voice artists to have a place in the future.

From Voice Actor to AI Voice

For decades, our job was simple.

A microphone.

A script.

A studio.

An actor.

A director.

Record.

Edit.

Deliver.

Today, the workflow is expanding.

Human recordings can become datasets.

Performances can become voice models.

Voice identities can become multilingual.

A Filipino voice recorded in Manila could potentially speak to audiences around the world.

One performer can become part of interactive education, conversational agents, games, localization, accessibility, audiobooks, digital humans, virtual characters, and entirely new entertainment formats.

The opportunity is enormous.

But only for people willing to understand the technology.

You cannot participate in a revolution that you refuse to study.

v4 Turbo and the Rise of Real-Time AI Voices

ElevenLabs is also positioning v4 for conversational agents.

Lower latency makes AI conversations feel more fluid, and the architecture allows speech generation to begin while the underlying language model is still generating its response. ElevenLabs says the system can also adapt its delivery during confrontations, escalations, holds, and other customer-service situations.

This is where AI voices become more than narration.

They become interactive.

Imagine Filipino AI teachers.

Filipino customer-service agents.

AI tourism guides.

AI companions for senior citizens.

Interactive historical personalities.

Filipino-language learning assistants.

Virtual broadcasters.

AI characters that understand not only our language but also how Filipinos naturally communicate.

That is where this technology is going.

And that is exactly why Filipinos need to participate now.

The Voice Industry Has Two Choices

We can spend the next several years asking AI companies to stop.

Or we can sit at the table and help build the technology responsibly.

I chose the second path.

Because technology rarely waits for permission.

When radio arrived, storytelling changed.

When television arrived, broadcasting changed.

When digital recording arrived, studios changed.

When the internet arrived, media changed.

When smartphones arrived, communication changed.

Now artificial intelligence is changing voice.

Again.

The question is no longer:

"Will AI enter the voice industry?"

It already has.

The better question is:

"Will Filipino voice artists be part of building its future?"

My answer has always been yes.

AI Should Not Kill the Human Voice

AI should amplify it.

Preserve it.

Translate it.

Extend it.

Make it accessible.

And allow voices from countries such as the Philippines to reach audiences that traditional production economics could never reach before.

But the human must remain at the center.

Behind every legitimate cloned voice should be a person who gave permission.

Behind every authentic dataset should be people who understand what they agreed to.

Behind every culturally accurate Filipino AI voice should be Filipinos.

That is the future I want to help build.

ElevenLabs v4 shows how quickly that future is arriving.

More emotion.

More control.

Faster cloning.

Lower latency.

More than 90 languages.

The technology keeps moving.

And I am glad I did not run away from it.

I gave AI my voice because I wanted the Filipino voice to be heard in the future.

Not replaced.

Not forgotten.

Heard.

And this is only the beginning.

Experience ElevenLabs

Want to hear what modern AI voice technology can already do?

Try ElevenLabs here for free:

https://try.elevenlabs.io/c8yd62n74m4x

Pocholo De Leon Gonzales
The VoiceMaster of the Philippines
AI VoiceMaster
Founder, CreatiVoices Productions

The future has a voice. Let us make sure the Filipino voice is part of it.