Your voice, speaking another language.

Tapra transcribes, splits speakers, translates in context, and clones each labeled voice back into the video.

Final rendering stays in this browser

Ready

Choose a video to begin.

Source language

Tapra Voice · ar-AR

Target language

Base OmniVoice · en

Translated captionsBurn the translated rows into the final video.

Keep original quietlyPreserves ambience, but may leave faint source speech.

Vocabulary hints (optional)Add only words expected in this video, separated by commas.

Speakers

Tick one voice, or keep multiple speakers labeled and cloned separately.

Detect how many people are talking.

Transcription—

0.150 TND / min

Translation—

0.150 TND / 1k chars

Cloned speech—

0.200 TND / 1k chars · clone 1.25×

Estimated job

—

Each step is charged only when it runs. Speaker labels bundled with transcription are free.

Ready0%