Localize a video end to end
Take a finished video and ship it in another language: separate the speakers, dub while keeping each original voice, then burn in subtitles.
Split the speakers onto their own tracks
Upload the video's audio. Every speaker comes back on a separate track — most accurate with up to five people. If several voices overlap at the same instant, that moment is kept in each speaker's track and reported rather than guessed at, so you always know what was and wasn't cleanly isolated. Skip this step entirely if your source is already single-speaker.
About Speech Separation →Dub into the target language
Upload the video and pick the target language. Each speaker is translated and re-voiced in a voice matched to the original, so the dub keeps the character of the source rather than flattening everyone into one narrator. The estimate shown before you start is how much of your plan the job uses — not how long it takes to render.
About Dubbing →Add subtitles
Generate subtitles in the target language and either download them as a file for your editor, or burn them into the video during the dub. Burning in is the right choice for social platforms that ignore subtitle tracks.
About Subtitle Creation →Review before you publish
Listen for names, numbers and product terms — these are what translation gets wrong most often. Fix them in the transcript and re-render rather than patching the audio by hand.
Ready to try it?
Create a free account and work through this guide with your own files.
