Audio to SRT
Audio to SRT converts uploaded speech into a source-language transcript and separate SRT, VTT and TXT files. It detects the spoken language automatically and accepts files up to 10 minutes and 200 MiB.
Run transcription1 credit per started 30 seconds. Quote before running; failed runs release credits. No free transcription preview.
Transcript
Automatically detect the spoken language, review timestamped segments, then download subtitles or plain text.
From uploaded speech to captions
Audio to SRT converts uploaded speech into a source-language transcript and separate SRT, VTT and TXT files. It detects the spoken language automatically and accepts files up to 10 minutes and 200 MiB. SRT and VTT carry segment start and end times. TXT contains the same recognized words without timestamps.
How to transcribe your audio
- Upload MP3, WAV or M4A, up to 10 minutes and 200 MiB. Sign in to use the workspace.
- Review the quote. A server media probe checks the duration without transcribing it. Transcription costs 1 credit per started 30 seconds, with a minimum of 1 credit.
- Run transcription after reviewing the price. Credits are reserved while it runs, charged on successful delivery and released on failure.
- Review and download the timestamped transcript as SRT or VTT, or download a plain-text TXT file.
A 30-second file costs 1 credit, a 31-second file costs 2 credits, and a 10-minute file costs 20 credits. There is no free transcription preview. Dubbing prices are separate.
Choose a file format
| Format | Contains | Use |
|---|---|---|
| SRT | Numbered segments and timestamps | Import captions into your video editor |
| VTT | Timed segments with a WebVTT header | Add captions to a web player |
| TXT | Recognized words without timestamps | Read, search and edit the transcript |
Review the result
Transcription uses automatic speech recognition. Check names, numbers and timings against the recording, especially for noisy or overlapping speech. Segment normalization can merge or omit very short tail fragments. The workspace does not offer transcript editing, speaker renaming or reassignment. Correct downloaded files in your editor before publishing.
Upload only media you have the right to use, with any required permission from the people recorded. Transcription reads speech; it does not extract text shown in video frames. It does not burn captions into a video, synthesize speech or translate the transcript.
Related tools
Use video to SRT for video files. Get plain text from video. For video transcripts, see video to text. To translate speech and create a new voice track, use the audio translator.
Frequently asked questions
Upload the file, check its duration and credit quote, then run transcription. Review the timestamped text and download SRT, VTT or TXT.
MP3, WAV or M4A files up to 10 minutes and 200 MiB per upload are supported. Longer files must be shortened before uploading.
1 credit per started 30 seconds, with a minimum of 1 credit. A 30-second file costs 1 credit; 31 seconds costs 2; 10 minutes costs 20. The duration quote appears before you run. Failed runs release reserved credits.
No. Transcription runs after you accept the duration quote. The free dubbing preview belongs to dubbing mode, which uses different pricing.
No. ASR detects the spoken language automatically. The transcript stays in that language; transcription mode does not translate it.
No. The workspace displays timestamped segments but has no speaker reassignment or transcript editor. Download the files to make corrections in your editor.
No. Transcription delivers separate SRT, VTT and TXT files. It does not generate a new video or voice track.