Transcripts
Turn a video into text you can actually use.
Turn an interview, lesson or voiceover into text you can review and reuse.
Free plan · no credit card
How it works
Add the recording
Choose a media file and select the language being spoken.
Correct the transcript
Replay unclear passages and verify names, numbers and quotations.
Choose TXT, SRT or VTT
Use TXT for further writing or a timed file for subtitles.
Start by deciding what the text is for. A quotation needs careful checking against the recording; show notes may need editing into headings; a subtitle track needs timing as well as words.
Speech recognition produces a draft. Background music, a distant microphone, acronyms and people talking together can introduce mistakes or omissions.
- Plain-text output
- Download the words without subtitle timestamps as TXT.
- Timed alternatives
- Use SRT or VTT when the next step is a video player or editor.
- Review against the recording
- Jump through the transcript and correct the wording before reuse.
- Language and script choice
- Choose the spoken language; Roman output is an Ultra feature, not translation.
From a rough transcript to something publishable
For an interview, check every name and direct quotation. For a lesson, check terminology and numbers. For meeting notes, review decisions and owners; the transcript alone does not identify what was agreed.
Do not treat a missed word as a deliberate omission by the speaker. Replay uncertain passages and ask for clarification if the exact wording matters.
What this workflow does not extract
A silent slide deck has no speech to transcribe. Text in a screenshot, chart labels and visual actions are not converted into the transcript. Add those details yourself when the audience needs them.
This workflow begins with a media file on your device, not a public video URL. The video stays on your device. Both Candy Lite and Candy Ultra send extracted audio for cloud transcription; sign-in and an internet connection are required.
Common questions
Can I get timestamps in TXT?
The TXT download contains words without timestamps. Download SRT or VTT for timed cues.
Can I translate the transcript into English?
The transcription workflow follows the spoken language. Romanisation changes script; it does not translate meaning into English.
Can I process a long lecture?
Creator supports videos up to 30 minutes and Pro up to 90. Check both the per-video limit and your remaining transcription allowance.
