Synthesia Dubbing 2.0: Lip Sync in 140+ Languages

Synthesia released Dubbing 2.0 on July 15, 2026. The update rebuilds its video dubbing system with frame-accurate lip sync that holds through fast cuts, voice cloning that actually reproduces emotion and accent, and translations that fit the original video timing instead of overrunning it. Synthesia says the output is now publishable without agencies or heavy post-production. Enterprise customer Merck Group is already using it.

What Did Synthesia Change in Dubbing 2.0?

Synthesia rebuilt four core systems: a lip-sync model that tracks mouth geometry through fast cuts and multi-speaker scenes, voice cloning that reproduces emotion and accent across 140+ languages, timing-aware translation that respects original segment lengths, and segment-level editing that regenerates individual segments without re-rendering the full video.

The previous version of Synthesia’s dubbing drifted out of sync during fast cuts and scene transitions, exactly the moments where a viewer’s eye catches a mismatch.

The lip-sync model now handles a specific challenge that plagued earlier AI dubbing: translating between languages with entirely different mouth shapes. English to Japanese, for example, required the model to map a completely different phoneme set onto the original speaker’s face. Dubbing 2.0 tracks the geometry of the mouth rather than mapping phonemes one-to-one.

Glossary support also works with dubbing for the first time. Product names, internal terminology, and regulated phrases stay consistent across every language version without someone fixing them by hand after each render.

Who Benefits Most from Dubbing 2.0?

Enterprise teams have dealt with a bad choice for years: dub fast and accept rough output, or wait weeks for polished localizations. Dubbing 2.0 is Synthesia’s attempt to remove that choice entirely.

The most obvious use case is L&D. Compliance training that used to stagger its rollout across languages (because each one needed fixing) can now go live everywhere on the same day. Corporate comms teams get something similar: a CEO message that actually sounds like the CEO in Japanese, not a flat robotic reading of translated text. Marketing gets simultaneous regional launches instead of waiting on the slowest language in the set.

Florian Metz, Global Head of Analytics and AI Product Portfolio at Merck Group, confirmed the impact: “Dubbing 2.0 has allowed us to scale our multilingual content in a way that was simply not possible before.”

Pricing and Credit Costs

Dubbing 2.0 is available on all Synthesia plans starting at $18 per month (annual billing). Lip-synced dubbing costs 240 credits per minute, dubbing without lip sync costs 120 credits per minute, and every user gets 450 free dubbing minutes through August 15, 2026.

The credit system determines how much dubbing you can do within your monthly allowance:

• With lip sync: 240 credits per minute of dubbed video
• Without lip sync: 120 credits per minute
• Starter plan: 1,200 credits/month ($29/mo or $18/mo annual)
• Creator plan: 3,600 credits/month ($89/mo or $64/mo annual)
• Enterprise: Unlimited video creation, credits for dubbing and API

Enterprise plans unlock the advanced editing workflow (transcript and translation refinement without burning credits on re-renders) and unlimited dubbing options. Self-serve plans can dub immediately but burn credits on each render pass.

How Does Synthesia Compare to Other AI Dubbing Tools?

Synthesia offers the widest language coverage (140+) and an integrated video platform, but Dubly.AI scores higher on raw lip-sync quality (96.4/100 vs. Synthesia’s newer model), Rask AI handles high-volume audio dubbing at lower cost, and ElevenLabs produces the best voice quality for audio-only dubbing without lip sync.

Dubly.AI only covers 38+ languages and starts at €79/month. If you need GDPR-compliant servers in Germany and the best raw lip sync on existing footage, Dubly is hard to beat.

Rask AI is cheaper and handles high-volume audio dubbing, but its lip sync scored 51.8/100 in the same benchmarks. Fine for e-learning where slides fill the screen. Less fine for talking-head content.

ElevenLabs Dubbing Studio has the best voice quality for audio-only dubbing across 32 languages, but it does not do lip sync at all.

Where Synthesia fits: teams already producing content on its platform (avatars, video creation, multilingual player) can now dub without exporting to a separate tool. The 140+ language count also outstrips every competitor listed above.

What Does This Mean for AI Video Translation?

The previous generation of AI dubbing tools still required enough human intervention that many teams kept their localization agencies on contract. If Synthesia’s first-pass quality claim holds up in practice, those agencies move from producing translations to spot-checking them.

The timing matters too. Google launched personal avatars in Google Vids one day later (July 16), Dubly.AI keeps winning lip-sync benchmarks, and ElevenLabs is pushing its own dubbing studio. Synthesia’s bet is that nobody wants to export video to a separate dubbing tool when the same platform already handles avatar creation and distribution.

FAQ

Sources

1. Synthesia - Introducing Dubbing 2.0 (July 15, 2026)
2. Synthesia LinkedIn announcement (July 15, 2026)
3. Synthesia AI Dubbing feature page
4. Synthesia Help Center - Credits for Enterprise