Sync (Sync Labs) Review: AI Lip-Sync API for Developers — Is It Worth It for Australians?

Verdict (TL;DR): Sync Labs is the leading developer API for AI lip-sync — technically impressive, well-documented, and genuinely useful if you're building a product. For a non-technical Australian business owner who wants to dub a video or put words in an avatar's mouth, it is the wrong tool entirely. The consent and legal risks of lip-syncing real people without authorisation are also squarely on the user.

Score: 13/25 — ⭐⭐⭐ Good

Last reviewed: 26 June 2026 by Adam.

What Sync is

Sync Labs (sync.so) is a San Francisco–based company that provides AI lip-sync technology as an API. The core product is not a video editor or an avatar studio — it is a programmatic service that accepts a video of a face and an audio file, then returns a new video where the face's mouth movements have been reanimated to match the audio.

The practical use cases range from dubbing content into other languages (so that on-screen talent appears to be speaking in the new language rather than being dubbed with mismatched mouth movement), to animating a still photo or video avatar to deliver any script, to integration into live video pipelines.

Sync Labs has released several model generations. As of June 2026, the main production models are lipsync-2, lipsync-2-pro, and sync-3, each representing different trade-offs between speed, quality, and cost. The API supports Python and TypeScript SDKs and is documented via an OpenAPI spec. There is also a lightweight web studio for non-API use, but the product is fundamentally developer-oriented.

The company raised a seed round in 2023 and positions itself alongside tools like HeyGen and Synthesia, while targeting the API-integration layer rather than the end-user video creation market.

What it costs in Australia

Sync Labs prices in USD. There is no AUD billing, and there is no GST applied at checkout — Australian business users should check their own GST obligations when purchasing offshore SaaS.

All plans include API access, the web studio, and the Python/TypeScript SDKs. The subscription tier unlocks higher limits and volume discounts; actual generation is billed per second of output video.

Usage pricing on top of subscription (USD, per second of output video):

AUD conversions use an indicative rate of 1 USD = 1.40 AUD as of September 2026. GST treatment of offshore digital services: under Australian law, non-resident suppliers may be required to register for and collect GST on B2C sales. Verify with your accountant. Exchange rate and GST status — verify at time of purchase.

A one-minute video processed through sync-3 would cost roughly USD $6.42–$7.98 in usage fees (≈ AUD $9–$11) plus whatever subscription plan tier you are on.

How it works in practice

You submit a job via the API: a video URL and an audio URL. Sync Labs processes the file on its servers and returns a new video. The web studio provides a drag-and-drop interface for the same workflow without writing code.

Processing is asynchronous — you poll a job endpoint for status. For a 30-second clip on lipsync-2-pro, typical turnaround is under two minutes. For sync-3, which performs diffusion-based super-resolution to enhance detail around teeth and facial features, processing takes longer but produces noticeably sharper results.

The SDK supports webhooks so your application can receive a notification when a job completes rather than polling.

Multi-language dubbing is supported: you provide audio in the target language and the tool reanimate the face to match. This is the use case that draws most interest from media and localisation teams — an English-speaking presenter can appear to address an audience in French, Spanish, or Mandarin without a re-shoot.

The free tier is genuinely useful for evaluation: three generations per month (each up to 20 seconds), one of which can use the top-tier sync-3 model.

What it does well

API quality. The documentation is thorough, the SDKs are well-maintained, and the job management interface is clean. For a developer building lip-sync into a product — a dubbing pipeline, an interactive avatar, a content localisation tool — Sync Labs is the easiest on-ramp to production-quality results.

Model range. The tiered model lineup means you can optimise cost vs. quality per use case. Cheap, fast lipsync-2 for high-volume social content; sync-3 for broadcast-quality close-ups where tooth and skin detail matters.

Free tier for testing. Three free generations per month, no credit card required, is genuinely useful. You can evaluate output quality before committing to spend.

4K support. sync-3 supports 4K output — relevant for post-production pipelines rather than social video, but meaningful for the segment of users building professional content tools.

What it doesn't do well

Not a no-code tool. There is a web studio, but Sync Labs is fundamentally built for developers. A small business owner who wants to create a talking avatar video for Instagram will not find this intuitive. They should look at HeyGen or Synthesia instead.

Cost uncertainty. The combination of a subscription fee plus per-second usage makes budgeting difficult without careful modelling. A content team producing a lot of long-form video could rack up meaningful usage charges that aren't immediately obvious from the plan page.

No avatar creation. Sync Labs does not create avatars from scratch. You need to provide a source video of a face. It reantimates what you give it — it doesn't generate a digital human from a text description or a single photo.

Data residency. Sync Labs is a US-based company. Video data submitted via the API is processed on US servers. There is no Australian or Asia-Pacific data residency option, and no public data processing agreement (DPA) specific to Australian privacy law obligations. Businesses handling sensitive footage — healthcare, legal, financial services — should assess this risk before integrating.

No output language detection. The tool reantimates lip movements to match audio — it doesn't generate or translate the audio itself. You need a separate TTS or translation layer to produce the dubbed audio before Sync can do its work.

Australian fit — likeness and voice consent

This is the section that matters most for Australian users.

Sync Labs' core capability — taking a video of a real person's face and making it appear they are saying something they never said — carries significant legal and reputational risk if used without the subject's explicit consent.

In Australia, the law in this area is developing fast:

• NSW amended the Crimes Act 1900 in February 2026 to criminalise the creation and distribution of non-consensual AI-generated sexual deepfakes.
• South Australia has laws covering invasive, humiliating or degrading AI-generated images of identifiable real people, with fines up to $20,000 and up to four years' imprisonment.
• An Australian MP has called for Denmark-style legislation granting people the right to own their likeness — which, if passed, would tighten the rules further.
• Outside sexual content, the legal landscape is less settled, but using someone's likeness without consent in commercial advertising or content could still attract liability under passing-off, personality rights, or consumer law frameworks.

The responsibility for consent sits entirely with the user, not with Sync Labs. The platform's terms of service prohibit creating deepfakes of real people without consent, but the platform cannot verify whether consent exists. Australian businesses using Sync Labs to lip-sync real people — employees, talent, public figures — must have documented consent before doing so.

For advertising specifically: the ACCC and Australian Consumer Law require that consumers are not misled. An AI-generated video that makes it appear a real person endorsed a product when they didn't could constitute misleading conduct. This is not hypothetical — it is the same principle that governs testimonial advertising.

Practical rule of thumb: use Sync Labs to animate faces where you own the rights (your own likeness, licensed stock talent, AI-generated avatars). Do not use it to animate real people without their documented, specific, voluntary consent.

Data: all video submitted to sync.so is processed on US servers. No AU data residency. (specific data retention period — check sync.so/privacy directly.)

Alternatives

• HeyGen (review-heygen-australia-2026-06-03) — the closest consumer-facing equivalent. HeyGen includes lip-sync as one feature inside a full avatar video platform, with a no-code interface. More accessible for non-developers; comparable quality for short-form content.
• Synthesia (review-synthesia-australia-2026-06-03) — AI avatar video with its own licensed presenter models. No lip-syncing of arbitrary faces; instead gives you a curated library of pre-consented avatars. Lower legal risk for business use.
• Colossyan (review-colossyan-australia-2026-06-26 — AWAITING REVIEW in Drafts) — AI presenter video with avatar creation. Positioned for learning and development content.

Bottom line

Sync Labs is a genuinely excellent product — if you are a developer building lip-sync functionality into a product or pipeline. The API is solid, the model range is impressive, and the free tier lets you evaluate quality with no commitment.

For everyone else — small business owners, marketers, content teams without engineering resources — this is not the tool. Use HeyGen or Synthesia, which give you similar output via a point-and-click interface.

And regardless of technical capability: every Australian business using AI lip-sync on real people needs explicit consent from the subject, full stop. The legal risk of getting this wrong is real and growing.

Score

Australian Fit gate: ⭐⭐⭐⭐⭐ Excellent requires AU Fit ≥ 4/5. Score: 2/5. Displayed rating capped at ⭐⭐⭐ accordingly. See Methodology.

Frequently asked questions

Can Australians use Sync Labs (sync.so)?

Yes, Sync Labs is accessible from Australia. Pricing is in USD only — there is no AUD billing option. All video data is processed on US servers, with no Australian data residency. Australian business users should verify their GST obligations for offshore digital services.

Is it legal to lip-sync a real person using Sync Labs in Australia?

Using Sync Labs to animate a real person's face without their consent carries legal and reputational risk in Australia. NSW and South Australia have laws targeting non-consensual AI-generated content. Commercial use of someone's likeness without consent could also breach Australian Consumer Law. Explicit, documented consent from the subject is required.

Do I need coding skills to use Sync Labs?

Sync Labs is primarily an API product aimed at developers. There is a basic web studio for non-technical use, but the platform's strengths — SDK integrations, webhook support, programmatic job management — require development skills. Non-developers wanting to create talking avatar videos should consider HeyGen or Synthesia instead.

Sources

• Sync Labs pricing page: https://sync.so/pricing — consulted 26 June 2026
• Sync Labs documentation — introduction: https://sync.so/docs/introduction — consulted 26 June 2026
• Sync Labs lip-sync models: https://sync.so/docs/models/lipsync — consulted 26 June 2026
• Sync Labs privacy policy: https://sync.so/privacy — consulted 26 June 2026
• NSW Government deepfake law reform: https://dcj.nsw.gov.au/news-and-media/media-releases/2025/nsw-government-strengthens-protections-against-deepfakes-and-ima.html — consulted 26 June 2026
• Gilbert + Tobin — Australian Government targets sexually explicit deepfakes: https://www.gtlaw.com.au/insights/australian-government-targets-sexually-explicit-deepfakes — consulted 26 June 2026
• AUD/USD conversion rate: indicative 1 USD = 1.40 AUD, September 2026

Wayback Machine snapshots: (capture pricing and privacy pages at time of publication per Verification Standard.)

The AI Directory Australia is an independent editorial publication. We don't accept payment for positive reviews. Read our Methodology and Affiliate Disclosure.