Eleven v3 Audio Tags: Emulating accents with precision

With Eleven v3 Audio Tags, switching accents is as simple as writing a bracketed cue. You can move between American, British, French, Australian — or any supported accent — mid-sentence, mid-script, or mid-character.

This opens new possibilities for creators who want dynamic, global, or expressive voice performances — without needing separate voice models or manual retakes.

What is accent emulation in AI speech?

Accent emulation is the ability to shift a voice’s pronunciation and rhythm to match different regions or dialects. It’s not a translation — the words stay the same — but the way they’re spoken changes.

With tags like [French accent], [Australian accent], or [Southern US accent], you can direct the model to speak in-region — and switch seamlessly when needed.

Example: [American accent] Could you switch my accent in the old model? [dismissive] Didn’t think so. [cheeky][Australian accent] But you can now — check this out, mate! [French accent] My love… eez like a red, red rose.

This isn’t imitation — it’s native delivery in context. The source voice you use will impact on the quality of the accent defined by an Audio Tag.

Why accent control matters

Accent emulation gives you creative and cultural range. You can localize content, define character identity, or make dialogue feel geographically grounded — all from a single model.

For example:

• An audiobook narrator who voices different characters from around the world.
• A game NPC who adjusts their accent based on location or player language.
• A product demo that speaks in local English dialects for different markets.

Accent cues let you script these experiences directly, without any model switching.

Common tags for accent and dialect

These tags help define regional identity and tone:

• Global English variants: [American accent], [British accent], [Australian accent], [Indian English], [Irish accent]
• Continental accents: [French accent], [German accent], [Spanish accent], [Italian accent], [Russian accent]
• Regional dialects & characters: [Southern US accent], [New York accent], [Scottish accent], [pirate voice], [robotic tone]

These tags can be used with emotional or delivery cues to create layered performance: [British accent][exasperated] You’re telling me *this* is the solution? Brilliant.[Southern US accent][calmly] Don’t worry now. We’ve got time.

Mixing accents and roles

Accent switching is especially powerful when combined with Character Performance and Multi-Character Dialogue.

For example:

• Narrator: [neutral tone] And then he turned to the stranger.
• Stranger: [thick Irish accent] Haven’t seen you around here, have I?
• Hero: [American accent] Just passing through.

Each speaker feels distinct — even though the same voice model delivers every line.

Directing identity, not just voice

With Eleven v3, accent becomes a design element. It’s part of your character’s personality, your story’s setting, or your product’s tone.

And with Audio Tags, you can shift that identity on command — reliably and with expressive control.

Selecting the right voice

Professional Voice Clones (PVCs) are currently not fully optimized for Eleven v3, resulting in potentially lower clone quality compared to earlier models. During this research preview stage it would be best to find an Instant Voice Clone (IVC) or designed voice for your project if you need to use v3 features. PVC optimization for v3 is coming in the near future.