AI lip sync re-syncs a person’s mouth movements in footage you already have to a new audio track. You need it when the video shows a real, identifiable person, when you are dubbing an ad you already shot, or when the read runs past 15 seconds. For new video built from scratch, native-audio models generate synced speech in a single step.
Most articles on this treat lip sync as a quality question. Pick the tool with the cleanest mouth movement, run your clip through it, ship. That framing was right two years ago and it is wrong now, because the models that generate video also generate the speech inside it.
The question worth answering is narrower. Given the ad you are trying to make, is the separate lip sync pass real work or dead weight? The answer turns on one thing, and it is not output quality.
Key Takeaways
The dividing line is likeness, not quality. If a real, identifiable person appears in the ad, you need a lip sync or dubbing tool. If the person is synthetic or non-specific, native-audio generation does the job in one step.
Sora 2 refuses the job by design. OpenAI’s API documentation states that real people, including public figures, cannot be generated, and that character uploads depicting human likeness are blocked by default (developers.openai.com, July 2026). That is a product boundary, not a limitation you can prompt around.
Veo 3.1 caps at 8 seconds. Google’s Gemini API documentation lists durations of 4, 6, or 8 seconds, with 8 required for 1080p and 4K (ai.google.dev, July 2026). A 30-second presenter read is structurally out of reach.
Dubbing is the strongest B2B case. YouTube reported averaging more than 6 million daily viewers watching at least 10 minutes of auto-dubbed content in December 2025 (blog.youtube, February 2026).
Disclosure is now law in two places. New York requires conspicuous disclosure of synthetic performers in ads as of 9 June 2026, at $1,000 for a first violation. EU AI Act Article 50 deployer obligations apply from 2 August 2026.
DesignerBox generates video with native audio. It does not lip sync your existing footage. Both facts matter for picking the right tool, and the second one is covered below.
What AI lip sync actually does
AI lip sync takes video you already have and re-animates the speaker’s mouth to match a different audio track. The original footage, framing, lighting, and camera movement stay. Only the mouth changes.
That is a different operation from generating a video. A generator starts from a prompt or an image and produces new frames. A lip sync tool starts from your frames and edits them. The distinction sounds academic until you need your founder’s face in the ad, at which point it decides everything.
Four products get called “lip sync” and only two of them do this:
| Type | What it takes in | Example use |
|---|---|---|
| True video-to-video lip sync | Your footage plus an audio track | Re-cut a spokesperson read without a reshoot |
| Dubbing | Your footage plus a target language | Run one ad across six markets |
| Avatar generation | A script, plus an avatar you picked or trained | Presenter video with no shoot at all |
| Performance transfer | A driving performance plus a character reference | Act a character you generated |
Only the first two operate on footage you already own. Avatar tools generate a new clip and hand it back. If you needed the original preserved, they have not solved your problem.
Starting from a still image rather than video is a third job again, with its own model constraints. We cover that separately in how to make a photo talk with AI.
The dividing line is likeness, not quality
Here is the fact that reorganises this whole category. OpenAI’s API documentation for Sora 2 states that real people, including public figures, cannot be generated, and that character uploads depicting human likeness are blocked by default (developers.openai.com, July 2026).
Google applies a softer version of the same constraint. Veo 3.1’s person-generation parameter is restricted for image-to-video and reference-image modes, with additional regional restrictions across the EU, UK, Switzerland and MENA (ai.google.dev, July 2026).
Read those together and the market splits cleanly. Native-audio generation owns new content featuring people who do not exist. Lip sync owns content featuring people who do. That is a consent and identity boundary the model vendors built on purpose, and no amount of output quality moves it.
Which means the decision is not “which tool is better.” It is “does a real person appear in this ad.”
What native-audio models can and cannot do
The video models in the DesignerBox catalog generate synchronised speech as part of the generation, with no second pass. Verified specs, from each provider’s own documentation:
| Model | Native audio | Max duration | Source |
|---|---|---|---|
| Veo 3.1 | Yes | 4, 6, or 8 seconds | ai.google.dev, July 2026 |
| Sora 2 Pro | Yes | 4, 8, or 12 seconds | developers.openai.com, July 2026 |
| Seedance 2.0 | Yes | 15 seconds, multi-shot | seed.bytedance.com, July 2026 |
The duration column is the honest constraint. Veo 3.1 tops out at 8 seconds, which covers a hook and nothing longer. Seedance 2.0’s 15 seconds is the most room any of them gives you.
Google is also candid about reliability on its own model page, noting that creating videos with natural and consistent spoken audio, particularly for shorter speech segments, remains an area of active development (deepmind.google, July 2026). Budget for regeneration.
One accuracy note worth carrying: Google’s Vertex AI model page and its Gemini API documentation describe Veo 3.1’s audio support differently. Check the current docs before you commit a campaign to a spec.
When you still need a dedicated lip sync tool
Four cases, and they are not edge cases:
The footage already exists. Native models generate. They do not retrofit. If you shot the ad last quarter and the offer changed, a generator cannot help you.
A real, identifiable person appears. Your founder, a paid actor, a licensed influencer. Sora blocks this outright. Veo constrains it.
You are localising. The performance exists and needs to survive translation. This is the strongest commercial case in the category, and the volume data supports it: YouTube reported averaging more than 6 million daily viewers watching at least 10 minutes of auto-dubbed content in December 2025, and creators uploading multi-language audio saw over 25% of watch time come from the video’s non-primary language (blog.youtube, September 2025 and February 2026).
The read runs long. Anything past 15 seconds exceeds every native-audio cap currently shipping.
The reference case for all of this is still Cadbury’s Diwali campaign in India, which licensed Shah Rukh Khan’s likeness and used lip sync technology to generate thousands of personalised retailer ads. It won a Cannes Titanium Lion and later the Creative Effectiveness Grand Prix. Note what made it work: a licensed likeness, used at scale, with consent at the centre.
What the tools cost
Prices below are monthly-billed rates verified in July 2026. Several vendors in this category display annual-billed rates by default, which understates the monthly figure by 20% to 45%, so check which toggle you are reading.
| Tool | Entry paid tier, billed monthly | Takes your uploaded video |
|---|---|---|
| ElevenLabs | $6 (Starter) | Yes, via its lip sync product |
| Synthesia | $19 (Starter) | Yes, via AI Dubbing |
| HeyGen | $29 (Creator) | Yes, via Video Translate and Lip Sync |
| Vozo | $29 (Creator) | Yes, standalone or with dubbing |
Two scope notes worth knowing before you buy. ElevenLabs’ help documentation states that lip sync is not part of its Dubbing product and is available separately through its Image and Video tools (help.elevenlabs.io, July 2026). Synthesia’s uploaded-footage lip sync is framed around dubbing and translation rather than same-language re-cuts (synthesia.io, July 2026).
Avatar-first tools sit in a different bucket. Creatify, Argil and Tavus all build a reusable avatar from footage you upload rather than re-syncing the original clip. Best fit if you want a repeatable presenter, outside scope if you need your existing footage preserved.
Where DesignerBox fits, and where it does not
DesignerBox is a campaign pipeline built on your product photo. It includes 13 image and video models across six providers, and the video models generate native audio. It does not include a lip sync tool that re-syncs footage you already shot. If that is the job, one of the tools above is the right buy.
What DesignerBox covers is the other half of the split: new video generated from scratch, with speech, from your actual product. Veo 3.1 and Sora 2 Pro both sit in the catalog, and Marketing Studio turns one product photo into the ad set around them.
Video is the most expensive operation on the platform, and it is priced per second of output. A Veo 3 clip with audio at 8 seconds runs 6,400 credits, which is more than the Premium tier’s entire monthly allocation of 2,500. Sora 2 at 720p and 8 seconds is 1,600 credits. Seedance Pro Fast at 720p and 5 seconds is 150.
AI video needs the Premium plan at $75 a month or higher. The full ladder is Free at 112 credits, Basic $15 for 500, Pro $35 for 1,000, Premium $75 for 2,500, and Ultra $200 for 8,000. Worth doing the arithmetic against your shot count before committing. Our breakdown of AI video generation cost runs the per-model math in detail.
Disclosure and consent
Lip-syncing a real person’s face is where this category meets the law, and the rules moved recently.
New York. General Business Law section 396-b, effective 9 June 2026, requires anyone producing an advertisement to conspicuously disclose that a synthetic performer appears in it, where they have actual knowledge. Penalties are $1,000 for a first violation and $5,000 for subsequent ones. Audio-only ads are exempt, as is AI used solely for language translation (nysenate.gov and governor.ny.gov, July 2026).
The EU. AI Act Article 50 deployer obligations apply from 2 August 2026. If you deploy a system that generates or manipulates video constituting a deepfake, you disclose that the content is artificially generated. The machine-readable watermarking duty sits on the model provider. The human-facing disclosure sits on you (artificialintelligenceact.eu, July 2026).
Right of publicity. California Civil Code section 3344 names voice explicitly and requires prior consent for advertising use, with a $750 damages floor plus profits and fees. New York Civil Rights Law sections 50 and 51 require written consent obtained first, and make violation a misdemeanour.
Platforms differ, and the direction surprises people. TikTok requires AI disclosure on all ads through a toggle in Ads Manager, and rejects or restricts undisclosed AI content (ads.tiktok.com, July 2026). Meta requires advertiser self-disclosure only for social issue, election and political ads, and from 1 June 2026 applies its own “AI Info” label to commercial ads through automated detection, with no advertiser action needed (transparency.meta.com, July 2026). Posting an AI video organically on Meta carries a disclosure duty that the same video running as an ad does not.
The practical standard for consent, drawn from the 2025 SAG-AFTRA Commercials Contract, is worth following even outside union work: written, separate from the employment contract, clear and conspicuous, and containing a reasonably specific description of the intended use.
FAQ
Can I lip sync a video with AI for free?
Several tools offer free tiers, generally watermarked and capped at a few minutes. For commercial ad use you will need a paid tier, since commercial licensing typically starts at the first paid plan. Entry paid pricing across the category runs roughly $6 to $29 a month billed monthly, verified July 2026.
Do I still need lip sync if I use Veo 3.1 or Sora 2 Pro?
Not for new content. Both generate synchronised speech as part of the generation. You need a separate lip sync tool when the footage already exists, when a real identifiable person appears, when you are dubbing into another language, or when the read exceeds 15 seconds.
Why can’t I use Sora 2 Pro to make a video of my founder?
OpenAI’s API documentation states that real people, including public figures, cannot be generated, and that character uploads depicting human likeness are blocked by default (developers.openai.com, July 2026). For a real person on camera, you need footage of them plus a lip sync or dubbing tool.
Does DesignerBox do lip sync?
No. DesignerBox generates video with native audio through models including Veo 3.1, Sora 2 Pro and Seedance 2.0. It does not re-sync mouth movements in footage you already shot. For that job, a dedicated dubbing or lip sync tool is the right tool.
Do I have to disclose AI lip sync in a paid ad?
It depends on the platform and where the ad runs. TikTok requires disclosure on all ads. Meta requires advertiser disclosure only for political and social issue ads. New York requires conspicuous disclosure of synthetic performers as of 9 June 2026, and EU AI Act Article 50 obligations apply from 2 August 2026.
What consent do I need to lip sync a real person?
Written consent, obtained before use, covering voice as well as likeness, and describing the intended use specifically. California and New York both require prior consent for commercial use, and New York requires it in writing. The SAG-AFTRA standard of a separate, conspicuous, specifically signed consent is the safest template.
Is dubbing the same as lip sync?
No. Dubbing replaces the audio track. Lip sync additionally re-animates the speaker’s mouth so it matches the new audio. Some tools do both, some only dub. ElevenLabs’ documentation, for instance, states that lip sync is not part of its Dubbing product and lives in a separate tool (help.elevenlabs.io, July 2026).
Where to go next
If you are generating new video ads from a product photo, start with the model catalog and the credit math. If you are re-syncing footage you already own, buy a dubbing tool and skip the generators entirely.
Related reading: AI video generation cost, how to make a photo talk with AI, how to create TikTok video ads with AI, AI image to video for ecommerce, and the best AI UGC ad tools.
Model capabilities verified against provider documentation from Google, OpenAI and ByteDance as of July 2026. Tool pricing verified from vendor pricing pages as of July 2026, monthly-billed rates unless stated. Legal and platform policy verified from primary sources including nysenate.gov, governor.ny.gov, artificialintelligenceact.eu, transparency.meta.com and ads.tiktok.com as of July 2026. Model specs and platform policy in this category change monthly. Verify before you commit a campaign.
