Upload and transcribe
Open dubbing and subtitles, upload a video or pick one from your assets, and press Transcribe.
Video transcriptionComing soon
Every spoken line, timed and ready to edit.
Upload a video and get what is said as a list of timed lines in the original language, then correct them and turn them into subtitles or a new voice-over.
Open the workspaceThis tool is in private testing and is not open to all accounts yet. The page describes how it will work; the workspace shows which tools your account can use today.
How it works
TrendRemix video transcription reads the speech in a video and returns it as timed lines in the language that is spoken. Each line is a sentence or natural phrase of up to about eight seconds, with start and end times in seconds checked against the file's measured length. Music and sound effects are ignored, and the detected language is recorded with the transcript. The lines open in the dubbing and subtitles editor, where you can correct words and timings, then export an SRT file, burn subtitles into the picture or re-voice the video. Transcription takes MP4, MOV or WebM files up to 18 MB that contain an audio track.
From input to finished result.
Correct the lines
Fix misheard words, move start and end times in tenths of a second, and add or delete lines.
Put the transcript to work
Keep the original sound and export an SRT, burn subtitles into the video, or choose a voice to re-voice the corrected lines.
What you can do
Timed lines, not a wall of text
Speech is split into sentences or short phrases with start and end times, the shape subtitles and voice-overs need.
Timings kept inside the clip
Times are checked against the measured length of the file and put in order, and overlaps and very short fragments are cleaned up.
Original language kept
Lines are written as they are spoken, without translation, and the detected language is noted.
Straight into subtitles or dubbing
The corrected lines feed the subtitle and voice-over options in the same editor.
Who it is for
- Subtitles for product videos
- Turn a spoken demo into subtitles for people who watch with the sound off.
- Starting a translation
- Get a clean source script to give a translator before re-voicing the video.
- Checking spoken claims
- Review exactly what a creator said about your product before you reuse their video.
At a glance
| Input | MP4, MOV or WebM with audio, up to 18 MB |
|---|---|
| Line length | One sentence or phrase, up to about 8 seconds |
| Timing | Start and end in seconds, within the file's length |
| Language | Written in the spoken language; language detected |
| Output | Editable timed lines, saved as a JSON transcript |
| Cost | Credits per video, returned if no speech is found or the task fails |
Before you download
- Check names, brand terms, prices and numbers, which are the words most often misheard.
- Play the video beside the lines and correct any start time that is early or late.
- Look for speech under loud music or from two people at once; these passages may be incomplete.
- Remove filler words if the lines will become subtitles.
Questions and answers
Which languages can it transcribe?
It writes down what is spoken in the language it is spoken in and records the language it detected. Check the result carefully for less common languages and strong accents.
Can I download an SRT file?
Yes, through dubbing and subtitles: keep the original sound and choose a separate SRT file. That creates a subtitle job from your corrected lines.
Why did I get no transcript?
The file had no audio track or no speech could be found. In both cases the task fails and its credits are returned.
My video is larger than 18 MB. What can I do?
Export a smaller or shorter version for transcription. Dubbing accepts files up to 200 MB, so you can still type lines by hand for the full video.
Does it translate the speech?
No. The transcript stays in the original language. Type a translation into the lines before re-voicing, or use the video translator.
Bring your product. Leave with something new.
Every result lands in your asset library, ready for the next tool.