Dubbing 2.0 is here: a new model, a new bar for quality

Today, we're releasing Dubbing 2.0, the biggest leap in dubbing quality since we launched the product.

Dubbing has always involved a trade-off. We're wired to sense when something's off, an unnatural tone, a distracting accent, lip sync that doesn't quite land, and in a business context that lack of authenticity is a real problem. When the goal is to inform and inspire people to act, content that "feels like AI" gets in the way.

So we improved dubbing to make it more human: a new state-of-the-art lip sync model, higher fidelity and more expressive voices, and translations that stay true to the original timing, all while preserving each speaker's original voice and delivery. Upgraded lip sync is now free and on by default, available on every plan.

Why we improved it

Dubbing has always been one of the most powerful ways to bring content to new markets, and AI has made it faster and more accessible than ever. As more teams used it for training, comms, and marketing at real scale, we saw an opportunity to push the quality further, tighter lip sync, more natural and expressive voices, and translations that stay truer to the original, so even more of your first pass is ready to go.

Dubbing 2.0 is that step change: the same workflow you know, now with output that keeps your message's meaning, professionalism, and human connection in every language.

What's new

• A new state-of-the-art lip sync model. Natural, studio-quality lip sync that holds alignment across fast cuts and transitions, and stays matched to the right on-screen speaker even when something briefly passes in front of the mouth.
• More natural, expressive voices. Higher fidelity audio with a wider emotional range and more natural pacing, preserving each speaker's original voice and delivery.
• Translations that keep their timing. Improved translation quality that stays close to the length and rhythm of the original, with Adaptive and Original timing modes to control pacing, so there's less to fix afterwards.
• A Glossary that carries your terminology. Brand names, product names, and acronyms apply automatically at translation time and stay consistent across the workspace, making terminology correction the exception rather than the rule.
• Frictionless transcript and translation editing. Review and refine both the transcript and the translation, and regenerate just the part you changed, with no credits used to review or edit. (Available on Enterprise.)
• From source to every market, faster. Record yourself, or upload any file, and we detect the source language automatically. Export per-language MP4s, WAV audio, or SRT and VTT subtitles, or share a single smart link that serves every viewer their language through one multilingual player.

What's next

This is just the start. We're already expanding our most expressive voices across more languages and pushing improved quality across every locale.

To learn more about how Dubbing 2.0 works, check out our documentation.