Timestamps for every subtitle line
Review each recognized line alongside its start and end time.
Turn speech from recordings and videos into timestamped subtitles you can download as SRT or JSON.
Up to 3 hours; media files up to 2 GB
MP4 / FLV / TS / AVI / MOV / WMV / MKV / MP3 / M4A / WAV
Account-protected files, processed in the cloud. Originals: 24 hours · Results: 7 days
Create subtitle drafts for interviews, meetings and videos, with timestamps that help you find each spoken line.
Review each recognized line alongside its start and end time.
Choose Simplified Chinese, US English, Spanish or Indonesian for transcription.
Your task continues after submission while you use other pages.
Use SRT with a subtitle editor or player, or JSON for structured text, timings and available speaker labels.
The quote uses the checked input duration. You see the total before starting.
Originals are available for 24 hours and transcripts for 7 days through your account.
Upload, select the spoken language, then download your transcript.
Choose supported audio or video up to 2 GB and 3 hours.
Choose the spoken language and whether to request speaker labels.
Check the timestamped transcript and download SRT or JSON.
Prepare subtitle files for interviews, recordings and video workflows.
Create timestamped text from recorded conversations for review and editing.

Turn supported recordings into subtitle lines you can review and export.

Generate an SRT or JSON transcript from supported video files.

1.2 credits per input minute. Minimum 3 credits per task.
Choose audio or videoFiles are checked after upload. Keep your original.
Each audio or video file can be up to 2 GB and 3 hours long. Longer recordings must be split before uploading.
Choose Simplified Chinese, US English, Spanish (Spain) or Indonesian to match the spoken language in your recording.
Download SRT subtitles for a player or editor, or JSON with structured text and timestamps. The tool creates separate files; it does not burn captions into a video. Review and edit the transcript in your usual subtitle editor.
Enable speaker labels to request a distinction between voices. Labels such as speaker_0 may be returned when available; they do not identify real people and may be incomplete or incorrect.
Noise, overlapping speakers, accents and unclear speech can cause missing or incorrect words. Check names, numbers and timings against the recording, then correct the downloaded file in a subtitle or text editor.
Check languages, transcript accuracy and how to use your downloads.