How to convert audio to text
Pick any audio or video file from your device. The upload is resumable, so a flaky connection will not lose your progress. Once the upload finishes, your recording joins the transcription queue and the page shows your position and an estimated wait.
When it is done, the transcript appears right on this page: every line carries a timestamp, and when more than one person is speaking, each line is labeled with the speaker. There is no engine to pick and no settings to tune. Every file runs on the same standard engine that powers the full product.
Accurate transcription, not a demo
This tool runs the exact same pipeline as the paid product, on infrastructure we operate. Your audio is not sent to third-party AI providers, and per-language accuracy is published openly on our benchmarks page. Speaker labels are included on every transcription, not sold as an add-on.
What the free tool includes
One transcription per day, files up to 45 minutes, speaker labels, timestamps, and automatic language detection across 52 languages and dialects. Free transcripts are deleted automatically about 24 hours after they finish, so copy what you need or sign up to keep them.
A free account adds signup credit, longer files, exports (TXT, SRT, VTT and more), a searchable library, and cited answers to questions about your recordings. See pricing for the full picture.
Common uses
Voice memos into notes, meetings into minutes, podcast episodes into show notes drafts, and lectures into study notes. If your recording is a conversation with several people, the free interview transcription tool leans on the same speaker labels to keep track of who said what. And if what you want to transcribe is already on YouTube, the free YouTube transcript generator takes a link instead of a file, and video files belong in the free video to text converter.