Video Translator

No. The video is opened by your browser and the audio track is extracted on your own machine. That audio is sent to be transcribed and the resulting text is sent to be translated, because neither can be done without a model. The video file, and everything visible in it, stays on your computer.

Three files if you want them: the translated subtitles as .srt and as .vtt, and the original-language transcript as .srt. No new video is produced and nothing is burned into the picture — that would mean re-encoding the whole file, which this tool does not do.

One page is one minute of audio, and the transcript is charged again as text when it is translated — roughly one extra page for every three minutes of speech. With 5 free pages a day that is about four minutes of video per day, start to finish.

Thirty minutes. Past that the browser is holding too much at once and the wait stops being reasonable. For a longer recording, cut the part you need with the Video Cutter first — that step runs entirely on your machine.

Clear single-speaker speech comes out well. Names, jargon, crosstalk, heavy accents and music underneath all cost accuracy, and the translation is machine translation on short lines with little context around them. The audio is sent in two-minute pieces, so a word landing exactly on a boundary can be lost. Read it as a good draft, not a transcript of record.

You get told that nothing was heard, and the pages are put back. Silence, music without vocals and unrecognisable noise all end the same way.