Auto Captions AI: Getting Caption Sync Right in 2026
Auto captions only look right when they are timed against the final audio of your video. Learn a practical Flixly workflow for AI voiceovers, lip-synced clips and burned-in captions that stay in step with the speech.
TL;DR
Flixly's Auto Captions tool listens to the audio in the video you upload, transcribes it and burns the captions into a new MP4. Sync problems usually come from captioning too early. Finish the voiceover, lip sync and edits first, export the final cut, then upload that file. Pick a caption style and position, check the quote, and review the result before you publish.
Common assumption about auto captions
Many people think auto captions just need to get the words right and the timing will sort itself out. In practice, most sync complaints have nothing to do with transcription. The captions were timed against one version of the audio, and the video was changed afterwards.
Auto captioning follows the audio track. If that track moves, gets trimmed, or gets replaced after captions are burned in, the text no longer lines up with the speech.
Why captions drift
These are the usual causes:
- Captioning before the edit is locked. Cutting a sentence, adding an intro or changing the pace after captioning shifts everything that follows.
- Swapping the voiceover. A new take or a re-generated voice has different pauses, even with the same script.
- Noisy or buried speech. Loud music or effects under the voice make it harder for the transcription to find where words start and end.
- Speed changes. Speeding a clip up or slowing it down after captioning breaks the alignment.
What to do instead
Treat captions as the last creative step. Build the video, lock the audio, export the final cut, and caption that exact file.
Auto Captions works on a video you upload. It transcribes the speech, burns the captions into the frame, and gives you a captioned video to download. There's no separate transcript or .srt to import, so the audio in your upload is what the captions follow.
Workflow for AI voiceovers
- Write the script and generate the voice in Text to Speech. Models include Gemini 3.1 Flash TTS, ElevenLabs Multilingual V2 and OpenAI TTS.
- If you need a specific voice, create it in Voice Cloning first.
- For a talking character, match the mouth to the audio with Lip Sync. Otherwise, place the voiceover on your footage in your editor.
- Finish every trim and transition, then export the final video.
- Upload that export to Auto Captions and pick a caption style and position.
- Check the credit quote, run it, and download the captioned video.
The quote depends on how long the video is, so short clips cost less than long ones.
Choosing a caption style
| Style | How it looks | Good for |
|---|---|---|
| Word by word | Words appear one at a time as they are spoken | Fast-paced short-form clips |
| Karaoke | Words are highlighted as they are spoken | Music, hooks and emphasis |
| Sentence | Full lines appear together | Tutorials, interviews, longer speech |
Position matters too. Bottom placement is familiar, but on vertical social video the lower part of the frame is often covered by app buttons and descriptions. Center or top placement keeps the text visible.
Verify the sync
Watch the captioned video from start to finish with sound on. Pay extra attention to the first line, any spot where you cut, and fast exchanges between speakers. If text consistently leads or trails the voice from one point onwards, something changed in the video at that point. Go back to the uncaptioned final cut and caption it again.
Common failure points
The most common mistake is captioning a draft and then editing it. Burned-in captions are part of the picture, so they can't be nudged later. Re-run the uncaptioned export instead.
The second is audio clarity. Lower background music under the voice before you export. Clear speech transcribes more cleanly and gives more reliable word timing.
Finally, if you re-generate a voice or switch to a cloned voice, even with the same script, treat it as a new video. The pauses and pacing change, so caption the new export rather than reusing an old captioned file.
Frequently Asked Questions
Can I upload an .srt file or a separate audio track to Auto Captions?▾
No. Auto Captions takes a video file, transcribes the speech in its audio and returns a new video with the captions burned in. Put your voiceover on the video before you upload it.
Does caption sync depend on a specific frame rate?▾
Auto Captions has no frame-rate setting. What matters is captioning the exact file you will publish, so the transcript is timed against that audio.
Can I fix out-of-sync captions after the video is rendered?▾
Burned-in captions can't be shifted afterwards. Run Auto Captions again on the uncaptioned final cut, especially if you trimmed or re-timed the video after the first pass.
Which caption styles are available?▾
You can choose word by word, karaoke or sentence captions, and place them at the bottom, center or top of the frame.
How many credits does auto captioning cost?▾
The cost depends on the length of your video. The app shows a quote once you choose the file, before you start the run.
