Correct an auto-generated transcript
Review suggested corrections against the original, accept or reject flagged suggestions, and preserve existing speaker and timestamp information.
Paste or upload a transcript. Get it back with clean punctuation, speaker labels, and formatting before delivery.
Upload a transcript or paste your text. Review suggested corrections, preserve the original meaning, and export a cleaner, client-ready document.
Supports TXT, DOCX, SRT, and VTT files up to 20 MB.
Review and correct existing text without re-transcribing the recording. Maximum 20 MB.
maya: yeah so we shipped the cut on thursday but the captions still overlap on the second line. jordan: right and the client needs the clean version by friday
Maya: Yeah, so we shipped the cut on Thursday, but the captions still overlap on the second line.
Jordan: Right, and the client needs the clean version by Friday.
Suggested fixes cover grammar, punctuation, capitalization, likely mishearings, times, and paragraph breaks. You accept or reject each one. Speaker labels and timestamps stay as they are.
Already have an auto-generated transcript or a draft from a transcription service? Upload it to edit misheard words, correct punctuation, manage speaker labels and prepare a clean client-ready export. You do not need to transcribe the recording again.
Review suggested corrections against the original, accept or reject flagged suggestions, and preserve existing speaker and timestamp information.
Rename or merge speaker labels and add elapsed timestamps or calculated SMPTE timecodes. These markers are not word-level audio synchronization.
Export TXT, DOCX or PDF. SRT/VTT exports require segments with actual start and end times; presentation timestamps alone do not create synchronized subtitles.
This editor fixes a transcript you already have. To create a new transcript from a recording, use Video to Transcript. For detailed client-specific formatting rules, use Client Guideline Format. Unlike audio and video editing software, changing words here does not edit the recording.
Before you upload: Check the current plan limits and pricing. Review any suggested changes before delivering a transcript to a client.
Review suggested fixes for grammar, punctuation, capitalization, likely mishearings, times, and paragraph breaks. Accept or reject each change. Speaker labels and timestamps stay put.
Rename, reassign or merge speakers across the whole transcript.
Add elapsed timestamps or calculated SMPTE timecode. It does not re-align words to audio.
Apply Rev-, GoTranscript-, TranscribeMe- or Scribie-style rules, or send the transcript on to Client Guidelines.
Your working copy is saved in this browser as you work.
TXT, DOCX, SRT or VTT up to 20 MB. Add the matching recording only if you need its duration or frame rate.
Edit any line, rename or merge speakers, with undo and redo.
Elapsed timestamps or calculated SMPTE timecode at your chosen interval.
TXT, DOCX or PDF. SRT and VTT when every segment has a start and end time, from an uploaded subtitle file or typed in.
Illustrative example: the meaning stays the same while punctuation, capitalization, paragraph breaks, and speaker formatting are cleaned up.
maya: yeah so we shipped the cut on thursday but the captions still overlap on the second line. jordan: right and the client needs the clean version by friday
Maya: Yeah, so we shipped the cut on Thursday, but the captions still overlap on the second line. Jordan: Right, and the client needs the clean version by Friday.
Correct an AI draft and deliver it in the client's format.
Clean interview transcripts and add timecodes for citation.
Consistent speaker labels and timecodes before review.
Text, speakers and timestamps are corrected on a transcript that already exists. Nothing is re-transcribed.
Yes. Upload the transcript you already have. The recording is optional and only used for duration or frame rate.
Upload the TXT, DOCX, SRT or VTT you already have, accept or reject the suggested fixes, rename speakers, and add elapsed timestamps or SMPTE timecode if you need them. Then export. The audio is never transcribed again.
TXT, DOCX, SRT or VTT, up to 20 MB.
Rename a speaker once to update every line, or reassign individual lines.
Yes: elapsed timestamps or calculated SMPTE timecode at an interval you choose.
No. Timecodes are calculated; words are not re-aligned to the recording.
Yes. Use the formatting presets here, or Client Guidelines for a full review workflow.
Yes, when every segment has a start and end time, from an uploaded SRT or VTT or typed in. Timecodes added on this page are labels and are not exported as subtitles.
Your draft is saved in this browser until you start a new transcript or clear site data. It is not synced to your account.
Fix text, speakers and timecodes, then export for delivery.
Upload Transcript to FixWhen the file already exists and the words are the problem.
Change labels on a file you already exported.
What those times mean, and when they are not speech alignment.