Transcription, translation and synthetic voices per character, with burned-in subtitles. Pick your target language from hundreds and generate your dub right here. Free text to speech and transcription, no account needed, in the same languages.
Not just dubbing: these tools work on their own, free of charge, with no account and no credit card.
Type up to 3,000 characters in one of 646 languages, describe or clone a voice and listen for free. HD download optional.
Transcribe a recording (MP3, WAV, M4A…) in over 1,000 languages: the plain-text transcript is free. SRT, JSON and translation optional.
Get the transcript of a video (MP4…) for free, language detected automatically, with SRT subtitles and translation as options.
Snowing AI dubs into hundreds of rare and under-resourced languages — for example Basque, Galician, Welsh, Kabyle, Kinyarwanda, Kurdish, Maltese, Swahili or Tibetan — directly online. New languages are added regularly.
You submit a video (YouTube URL or MP4 file). The AI transcribes the soundtrack, translates it, generates a voice per character and burns in subtitles. You watch the result right on the website.
A preview of the first 2 minutes is free. The full version is billed pro rata, from €2/min, online and without creating an account.
Yes. Our API (pipeline.snowing.ai) lets you submit a URL or a file, poll the job and download the result automatically.
No. Results are deleted automatically 4 hours after processing finishes.
Without an account or credit card: listening to and downloading text to speech in reduced quality, the plain-text transcript of an audio or video file, and a 15-second dubbed preview (2 minutes with a free email account). You only pay for extras: HD audio, formatted transcript with SRT subtitles, JSON and translation, or the full dub.
Yes. Text to speech, audio to text and video to text are standalone tools, available from the toolbar: turn up to 3,000 characters into speech with a described or cloned voice, or upload an audio or video file to get its transcript (up to 30 minutes per file, language detected automatically).