How Does the Video Translation Workflow Work, From Upload to Publish?

Translating a video ad in GeckoDub is a five-stage loop: prepare the file, upload and configure, proofread the translated script, dub, and publish. The only stage that needs your judgment is the proofreading step — the rest is upload, wait, download.

1. Prepare the source

Start from the highest-quality master you have rather than a platform re-download, and keep the file under the 500MB upload limit. Clear audio and a well-lit, unobstructed face give both voice cloning and lip-sync the most to work with.

2. Set your rules once

Before your first project, save your Custom Glossary (product names on the Don't Translate list, preferred wording per language) and your Brand Kit subtitle styling. These are saved presets — every later project follows them automatically, so this is a one-time setup, not a per-video step.

3. Upload and configure

Upload the video, or use bulk upload for a batch, and pick your target language or languages from the 70+ supported. Choose the features you want on this render: voice cloning, lip-sync (standard or HQ, which costs double the tokens), and animated subtitles.

4. Proofread before dubbing

GeckoDub transcribes and translates, then shows the original and the translation side by side. Read the hook, the offer and the CTA, edit anything that reads as translated, and watch the fit score so lines don't run past the original timing. Then click Apply Changes and Dub — for short ads, the render usually takes one to three minutes.

5. Review, download, publish

Preview the result: check lip-sync on close-ups, voice delivery, and subtitle placement against each platform's safe zones. Download, upload to Meta, TikTok or YouTube, set targeting for the new market, and launch.