Translate video speech into written text

How to translate a video or audio recording

Start with recorded audio or a local video. For video, Hushscript prepares the audio track in your browser and sends only prepared audio for online speech recognition; the source video stays on your device. The public preview transcribes up to 5 minutes in the original language and offers no translation controls. In normal mode, choose the written target languages before full transcription starts. Each translation arrives beside the original transcript for review and export. Audio translation follows the same process: an audio-only recording gets its written translation beside the original transcript.

Languages you can translate into

Choose from exactly 13 targets before full transcription: English, Spanish, French, German, Italian, Portuguese, Dutch, Hindi, Japanese, Chinese, Thai, Turkish, and Vietnamese. A Spanish recording can produce the original transcript in Spanish and an optional written English translation; for English to Spanish, recorded English audio produces written, translated Spanish text. English to Hindi works the same way and arrives as text, not a voice track.

This target list is smaller than, and separate from, the roughly 99 spoken languages Hushscript can transcribe. A supported source language is not automatically a translation target: if the language you want is absent from the 13 above, Hushscript cannot generate that translated output. If you are deciding whether you need transcription or translation, what “translate audio to text” means explains the difference.

	Source language	Target languages
How many	Roughly 99	Exactly 13
When set	Detected automatically in preview; automatic or manual in the app	Selected before full transcription
What you get	The original transcript	A written translated transcript per target

Saved dictionaries, keyterms, and prompts can help recognition produce a better source transcript, but they are not translation glossaries. They do not force names, jargon, or preferred wording into the translated text. Review each translated language tab before exporting it.

Translate video into text and subtitle files

After reviewing a written translation, export it in any of the same 21 transcript formats as the original transcript. Subtitle choices are SRT, VTT, ASS, TTML, SBV, and LRC. FCPXML and EDL are editor-timeline formats, and the password-protected .husharchive is a separate bundle rather than another export format.

Hushscript returns written transcript and subtitle files. It does not synthesize a translated voice, dub the video, replace its audio track, or translate a live microphone or stream. Import a translated subtitle file into a compatible player or editor when you want captions alongside the source video.

Translation cost and mode limits

Translation is optional. Base transcription uses the recording’s duration, and each selected target adds 25% of that duration to billed minutes. A 60-minute recording translated into two languages uses 60 base minutes plus 30 translation minutes, for 90 minutes total. The same calculation applies at any length:

Recording	Target languages	Billed minutes
60 min	2	60 + 30 = 90
12 min	1	12 + 3 = 15
12 min	3	12 + 9 = 21

Transcribing only the original language remains available with no translation surcharge. Translation runs only in normal mode; private mode saves no transcript to the account and rejects translation. Hushscript encrypts stored translated transcript content at rest, removes its transcription-audio copy after processing, and never trains on recordings.

Updated: October 14, 2026