News · Meta adds Hindi and Portuguese to its Reels translation and lip-sync feature

Oct, 94 min to read
AI Products

Meta adds Hindi and Portuguese to its Reels translation and lip-sync feature

Meta AI now dubs and mouth-syncs Reels across four languages, gated by follower thresholds on Facebook and open to public Instagram accounts.

What shipped: two languages layered onto an existing pipeline

Meta added Hindi and Portuguese to its Reels translation feature on October 9, 2025, bringing the supported set to English, Spanish, Hindi, and Portuguese. This is an incremental step on a pipeline that already existed: in August, Meta had extended bidirectional English–Spanish translation to Facebook and Instagram creators.

The mechanic is consistent across all four languages. Meta AI translates the spoken audio, dubs it in a voice meant to match the creator, and — if the creator opts in — syncs the translated audio to their mouth movements. The company frames Hindi and Portuguese as entries into 'some of the largest Reels markets,' which is the clearest signal of why these two were chosen next.

Voice cloning and lip sync are the actual product, not just translation

The interesting engineering claim is not that Meta translates text — it's that the output is meant to sound like the original creator. Per the announcement, Meta AI 'mimics the sound and tone of a creator's voice,' so the dubbed reel is intended to feel authentically like the person who filmed it.

A creator can also choose to enable the lip syncing feature, which syncs the translated audio to the creator's mouth's movements, creating a more natural viewing experience.Montana Labs

That combination — voice matching plus mouth-movement alignment — is what separates this from a captions or subtitle feature. It also makes lip sync an explicit opt-in rather than a default, which suggests Meta is treating the more manipulative-looking transformation as something creators should choose deliberately.

The access rules define who is actually in the rollout

The eligibility gates are specific and worth reading closely. On Facebook, translation is available to creators with 1,000 followers or more. On Instagram, it's open to all public accounts. Both are conditioned on being in a country where Meta AI is available.

The asymmetry — a follower threshold on Facebook, none on public Instagram accounts — implies Meta is scoping the feature differently by platform rather than shipping one uniform policy. And 'countries where Meta AI is available' is the real boundary: the language list expands faster than the geographic footprint of the underlying assistant.

Labeling and viewer opt-out signal how Meta wants synthetic dubbing perceived

Every translated reel carries a 'Translated with Meta AI' label, and viewers can turn translations off entirely or select 'Don't translate' to keep a reel in its original language. Those controls sit in the audio and language section of the three-dot menu.

For a feature that clones a creator's voice and moves their lips, disclosure is not a nice-to-have — it's how Meta positions AI dubbing as transparent rather than deceptive. The label plus a viewer-side off switch is the company's answer to the obvious concern: that audiences might not know a video was never spoken in the language they're hearing.

The specific implication: language coverage becomes a distribution lever

By making translation free and tying its language list to Meta's largest Reels markets, Meta is turning language coverage into a distribution mechanic rather than a premium feature. A creator filming in Hindi can now reach English, Spanish, and Portuguese audiences without re-recording anything.

For teams building on or around this ecosystem, the operative constraint isn't the model quality — it's the intersection of the four supported languages, the follower gates, and the countries where Meta AI is live. That intersection, not the promise of 'more languages coming soon,' determines who can actually use this today.

Find this story relevant to you?

Contact us to find a unique solution

Contact us

Need an AI engineering partner that can actually build?

We help businesses integrate AI, build AI-powered products, automate high-value workflows, and modernize the software systems behind them.

Get in touch

Related reading

More analysis around product delivery, operational AI, and the systems work that makes deployment hold up in reality.

Jul, 144 min to read
AI Products

How Google DeepMind rebuilt Pelé's unfilmed 1959 goal from archives and stunt footage

Jul, 134 min to read
AI Products

Expedia's image-selection automation is the concrete piece behind its AI marketing story

Jul, 134 min to read
AI Products

ENEOS Materials built over 1,000 custom GPTs and put ChatGPT Enterprise in front of every employee