Import subtitles with timecodes
SRT and VTT are supported. Each subtitle becomes a section with its start and end time, which you can also edit in the script view.
AI subtitle dubbing
Turn subtitles into voiceover that stays on time
Upload an SRT or VTT file and every line plays at its original time. See at a glance which lines run past their subtitle, fit them back with one click, then export a full audio track that lines up with your video, plus the updated subtitles.
No video upload needed, just an SRT or VTT file
The hard part of dubbing a video isn't generating the voice, it's timing. You generate clips line by line, drag each one into your editor, and redo the alignment whenever a line changes. AI subtitle dubbing starts from your subtitle file instead. Import an SRT or VTT and each subtitle becomes a section with its original start and end time.
After generation, every section shows its subtitle window and the actual audio length. Lines that run long are flagged, and one click fits them back with playback speed, with no regeneration and no extra credits. If a line is far too long, you'll be asked to shorten it. When a line overruns, the following lines are pushed back rather than overlapping or being cut off.
When you're done, export a full MP3 where each line starts at its subtitle time and silence fills the gaps. Drop it at 0:00 on your video timeline and it lines up. You can also export SRT or VTT subtitles built from your edited lines and the real voiceover timing, plus a ZIP of section audio.
- Import with timecodes kept
- SRT / VTT
- One-click fit, no credits
- 1.0–2.0x
- Subtitle sections per project
- 200
- Timeline-aligned export
- MP3 / SRT / VTT
A voice for every speaker
Start a line with "Host:" or "Narrator:" to assign speakers, then pick a cloned voice or a public voice from the Voice Square for each.
See which lines run long
Each section shows its subtitle window, audio length, and how far it overruns or is delayed. A summary lists long, failed, and ungenerated sections so you can jump straight to them.
One-click fit
Speed up each long line just enough to fit its window, with no regeneration and no credits. If a regenerated line no longer needs the speed-up, you'll be prompted to lower it.
Preview on the timeline
Continuous playback uses the same gaps as the exported audio. If something sounds off, one click takes you to that line.
Export audio and new subtitles
Get a full MP3 aligned to the subtitle timeline, updated SRT and VTT files, and a ZIP of section MP3s.
Promos and short videos
Swap the voice or rewrite a few lines and re-dub on the same timing, without re-editing the footage.
Courses and screen recordings
Replace a recorded narration with a clear, steady voice, and drop the slips and filler along the way.
Other language versions
Import translated subtitles and use the same set of voices to produce an English, Japanese, or other language version.
Animation and game temp tracks
Rough out pacing with AI voices before the recording session, then hand final timing and lines to your voice actors.
- 1
Import subtitles
Upload an SRT or VTT file. Each subtitle becomes a section with its timecode, and you can review speakers and sections in the preview.
- 2
Assign voices and generate
Choose a voice for each speaker and see the credit cost before you start. Try a line or two before generating everything.
- 3
Check and fit long lines
See which lines run past their subtitle and fit them with one click. Shorten lines that are far too long and regenerate just those.
- 4
Export audio and subtitles
Export the timeline-aligned MP3 and place it at 0:00 on your video. Export new SRT or VTT files if you need them.
- 01
One sentence per subtitle
Subtitles of 2–6 seconds holding one complete sentence are the easiest to fit. Several sentences in one subtitle will almost always run long.
- 02
Budget words by speed
English voiceover usually runs about 2–3 words per second. A 2-second subtitle with 10 words will almost certainly run long.
- 03
Set the overall speed first
Try a line or two. If most run long, raise the speed in project settings before generating everything. It sounds more natural than speeding up lines one by one.
- 04
Keep speed-ups under 1.3x
Lines fitted at a high speed start to sound rushed. Trim a few words and regenerate those instead.
- 05
Use consistent speaker names
Use the same name for a speaker across every subtitle, for example always "Narrator", so all their lines get the same voice.
- 06
Place the track at 0:00
The exported MP3 starts at zero with silence before the first line, so it lines up with the start of your video.
| AI subtitle dubbing | Line by line + manual alignment | |
|---|---|---|
| Alignment | Each line placed at its subtitle start automatically | Drag every clip into place in your editor |
| Changing a line | Regenerate just that line, it stays in place | Regenerate, then realign by hand |
| Lines that run long | Flagged automatically, fit with one click | Found on playback, trimmed or sped up by hand |
| Subtitles | Export updated SRT / VTT | Edit the subtitle file by hand |
| Deliverables | Aligned MP3, section audio, new subtitles | A folder of loose audio clips |
- Import formats
- SRT, VTT (up to 10 MB), plus other script formats
- Sections per project
- 100–3,000 by plan
- Characters per section
- 500 (1,000 on Premium). Longer lines are split and their time is shared by length
- Speakers
- Free 2, Starter 4, Standard 8, Premium 20
- Timeline rule
- Each line starts at its subtitle time. If the previous line runs long, it's pushed back, never overlapped or cut
- One-click fit
- 1.0–2.0x playback speed, no regeneration, no credits
- Export formats
- Timeline-aligned MP3, section MP3 ZIP, SRT, VTT, TXT script
- Pricing
- 1 credit per Chinese character or 3 English characters, only for sections you generate
No. You only need an SRT or VTT file. The exported MP3 is aligned to the subtitle timeline, so place it at 0:00 on your video in any editor. We don't create or process video files.
Voice projects
Section-by-section voiceover for long scripts, audiobooks, and multi-speaker scripts.
AI voice cloning
Clone your voice from a 10–30 second recording and dub with it.
Voice Square
Find public voices for promos, tutorials, and animation.
AI podcast generator
Turn a script into a two-host podcast and export the full episode as MP3.
Try it with a subtitle file
Hear a voiceover aligned to your subtitles in minutes, then export the audio and subtitles when you're happy.