100+ languages, one transcript engine.
Auto-detected for every upload. Same $2 per hour, same speed, same speaker labels, same in-browser editor. No language toggle to remember.
Most common
The rest of the 100
What about your language?
Do I need to pick a language before uploading?
No. The model detects the spoken language automatically. If a file mixes two languages, it transcribes both — the output keeps each phrase in the language it was spoken in.
How accurate is non-English?
Near-human on clear audio for the top languages above. Other languages range from very good to usable depending on data coverage, accent, and audio quality. You can edit any line in the browser if a phrase comes out wrong.
My language isn't listed. Will it work?
The model supports more dialects and lesser-used languages than we explicitly list. Upload a short file to test — the $2 minimum covers anything up to one hour, so you can check accuracy on a real recording before committing.
Does the price change for non-English?
No. $2 per hour for every language, every file.
$2 per hour. No subscription.