> ## Documentation Index
> Fetch the complete documentation index at: https://docs.tryvinci.com/llms.txt
> Use this file to discover all available pages before exploring further.

# One ad, five languages

> Your best ad only speaks one language. Here is how to localise it without a reshoot, while keeping the brand sounding like itself.

Your best ad only speaks one language. Here is how to localise it without a reshoot, while keeping the brand sounding like itself.

> [card] What you get
>
> The same ad localised for five markets.

> [card] Time to read
>
> About 4 minutes. Each step works with any toolset.

## What you need

* A finished master ad
* A native speaker per language for the script
* A voice model, and lip-sync for on-camera shots

## Steps

**1. Lock the master cut**

Finish the ad in one language first. Every change after this point has to be made five times.

**2. Translate for speech, not for print**

Have each script adapted by someone who speaks the language, so lines fit the same timing and sound natural out loud.

**3. Voice each version**

Generate the voiceover per language with the same voice, or a consented clone of your presenter, so the brand sounds consistent.

**4. Lip-sync the on-camera shots**

Where someone speaks on screen, run lip-sync so mouth movement matches the new audio.

**5. Swap on-screen text and captions**

Replace every caption, title and end card per language, and check that longer words still fit the layout.

> [!TIP]
>
> **Leave room in the timing.** Some languages run longer. Leave a beat of breathing room in the master so translated lines fit.

> [!WARNING]
>
> **Avoid this: translating the captions only.** Subtitles alone leave the voice in the original language. If the ad is voice-led, localise the voice too.

## Before you publish

* [ ] Lock the master cut
* [ ] Translate for speech, not for print
* [ ] Voice each version
* [ ] Lip-sync the on-camera shots
* [ ] Swap on-screen text and captions

## Two ways to go further

> [card] [Let Vinci make it](https://tryvinci.com/meet?utm_source=guide\&utm_medium=docs\&utm_campaign=one-ad-five-languages)
>
> Send us the master ad and target markets and we deliver every language version, reviewed by native speakers. Book a demo to get started.

> [card] [Free and open source Vinci](https://github.com/tryvinci/vinci)
>
> Run the Vinci studio on your own machine with your own model keys. It shows the price before every job.

### Run this workflow in the open source studio

These skills cover this guide: `@tryvinci/tts`, `@tryvinci/voice-clone`, `@tryvinci/lipsync`, `@tryvinci/captions-burn`.

> [tab] Studio
>
> **1. Install and start in demo mode**
>
> ```bash theme={"system"}
> git clone https://github.com/tryvinci/vinci && cd vinci
> python3 -m venv .venv && . .venv/bin/activate && make install
> ./scripts/vinci init --demo
> ./scripts/vinci studio
> ```
>
> Demo mode uses practice providers, so it costs nothing and needs no key.
>
> **2. Open a skill**
>
> Open the studio at `http://127.0.0.1:4310/studio` and pick `tts`.
>
> **3. Connect your own keys when ready**
>
> Add provider keys under **Connect**. Live spend stays off until you set `VINCI_ALLOW_LIVE_SPEND=1`, and every job shows its price first.

> [tab] HTTP
>
> With the studio running, every skill is also available over HTTP:
>
> ```bash theme={"system"}
> curl -sS -X POST http://127.0.0.1:4310/runs \
> -H 'content-type: application/json' \
> -d '{"skill":"@tryvinci/tts","inputs":{}}'
> ```
>
> The studio lists each skill's inputs.

## More guides

> [card] [Turn one podcast into a week of shorts](https://docs.tryvinci.com/learn/podcast-to-shorts.md)

> [card] [Stop paying for AI clips you throw away](https://docs.tryvinci.com/learn/ai-video-cost.md)

> [card] [Make a product ad from one photo](https://docs.tryvinci.com/learn/product-ad-from-one-photo.md)

> [card] [Ten fresh ads from one winner](https://docs.tryvinci.com/learn/ten-ads-from-one-winner.md)


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.