Turn Speech
Into Text
Upload audio or video up to 30 minutes and get the transcript as text, plus an SRT file for captions. Choose one of 16 languages or let the tool detect it. The price shows before you run.
For interviews, lectures, meetings and short videos
Words on the page, ready to use
One run gives you a transcript to read and a subtitle file to upload.
Transcript as a text file
The full text of the recording, saved as a file in your gallery. Copy it, download it, or paste it into your notes.
SRT subtitles included
Every run also makes an SRT file with timings, ready for YouTube, Facebook, LinkedIn and most video editors.
16 languages
Pick the spoken language, or leave it on automatic and let the tool detect it from the audio.
Audio or video
Upload a recording, a meeting export or a video. Files can run up to 30 minutes.
Clean first for better text
A noisy file is harder to transcribe. Run the Cleanup tab first when the recording has hiss, hum or echo.
Pay for finished files only
The price follows the length of the recording and shows before you run. A failed run is refunded automatically.
Transcription in three steps
Upload the recording, pick a language, get the text.
Create a free account
Free to start. The price is shown before every run.
Upload and choose a language
Add an audio or video file up to 30 minutes long and choose the spoken language, or leave it on automatic.
Read, copy, download
Open the transcript in the result card, copy it, and download the text file and the SRT file.
AI Audio Transcription FAQ
What do I get from a transcription run?+
Two files. The first is a plain text transcript of everything said. The second is an SRT subtitle file, which splits the same words into timed captions. Both are saved in your gallery, and the transcript is shown in the result card so you can read and copy it right away.
Which languages are supported?+
16 languages, the same list as text to speech, plus automatic detection if you do not choose one. For the best result pick the language that is spoken for most of the recording. Recordings that switch between languages can come out less accurate.
How accurate is it?+
Clear speech with little background noise is transcribed very well. Accuracy drops with heavy accents, crosstalk, music under the voice and poor microphones, and names or technical terms may be spelled wrongly. Treat the result as a strong first draft and read it once before you publish it.
How long can the recording be?+
Up to 30 minutes, as audio or as a video file, within the upload limit shown in the form. The tool measures the file on our side, so the price on the run button matches what is charged.
How much does it cost?+
The price depends on the length of the recording, in a few steps, and shows before you run. Credit packs are listed on the pricing page.
Do I need a subscription?+
Pay once with credit packs, or go monthly. Pack credits never expire. Failed runs are refunded. You can open an account for free and look at the tool before you buy credits.