Turn audio and video into subtitles you can actually edit
CaptionsThis transcribes your recording and gives every word its own timestamp, so the subtitles you get back are editable rather than fixed. Change where a line breaks, split a long cue in two, or retype a misheard name — the timing follows, because it is anchored to the audio and not interpolated from character counts.
Export SRT when the captions need to travel — YouTube, Premiere, DaVinci Resolve, a client. Export FCPXML when you are finishing in Final Cut Pro and want titles that land on the right frame with your styling already applied.
How it works
- 1Drop in your file. Audio or video — MP3, M4A, WAV, MP4, MOV, WEBM. Video is decoded to audio in your browser, so a large file never has to upload in full.
- 2Transcription and alignment. Whisper produces the transcript, then a forced-alignment pass locates every individual word in the waveform, so each one carries its own timestamp.
- 3Edit the lines. Set words per line or break on natural sentences. Split, join, and retype lines — the timing stays correct because every word is anchored.
- 4Export. Download SRT for anywhere, or FCPXML at your project's frame rate for Final Cut Pro, with per-word styling preserved.
What you get
- Word-level timingEvery word carries its own timestamp, so splitting a line never drifts out of sync.
- SRT and FCPXMLOne universal format, one that carries fonts and per-word styling into Final Cut Pro.
- Frame-accurate exportFCPXML uses rational timecode at your sequence's frame rate — no drift over a long cut.
- Translation built inTranslate the transcript into 10+ languages while keeping the original timing.
- Edits that persistYour session is saved locally as you work, with undo, redo, and recent history.
- No subscription15 free minutes every month, then one-time credit packs that never expire.
Frequently asked questions
- Is CaptionsThis free?
- Yes. You can transcribe a clip up to 5 minutes without an account to see how it works, though downloading the subtitle file requires signing in. A free account gives you 15 minutes and 3 transcriptions every month. After that, credit packs are one-time purchases that never expire — there is no subscription.
- What file formats can I upload?
- Audio: MP3, M4A, WAV, OGG, and WEBM. Video: MP4, MOV, and WEBM. Video files are decoded to audio in your browser before upload, so a large video does not need to be uploaded in full.
- What is the difference between SRT and FCPXML?
- SRT is a plain list of timecoded text cues that works in almost every player and editor, but carries no styling. FCPXML is Final Cut Pro's interchange format: subtitles arrive as editable title clips with fonts, positioning, and per-word styling intact, at your project's exact frame rate.
- How accurate is the timing?
- Transcription runs on Whisper large-v3, then a separate forced-alignment pass locates each individual word in the audio. That means every word has its own start and end time, so you can split or join subtitle lines without the second half drifting out of sync.
- Can I translate the subtitles?
- Yes. Pick a target language and the transcript is translated segment by segment, keeping the original timing boundaries. You can export the translated track as SRT or FCPXML like any other.
- Is there a file size limit?
- Uploads are capped at 25 MB. Because video is converted to compressed audio in your browser first, that limit covers roughly two hours of speech from a typical recording.
- Do I need an account to download the subtitles?
- Yes. Transcribing a short clip works without one, but exporting the SRT or FCPXML file requires a free account. Signing in also gives you 15 minutes and 3 transcriptions each month.
- Do you keep my audio?
- No. Audio is sent to the transcription service, processed, and returned as text. It is not stored. Your edits are saved only in your own browser's local storage.