Faceless channels
Captions for faceless video, where the text is the show.
Make the narration easy to follow without covering the footage it talks about.
Start captioning freeFree plan · no credit card
Faceless video captions follow the narration, and the main decision is how much of the frame the words take. Full-frame typography suits a text-led section, and smaller subtitles suit footage viewers need to see.
Video stays on your device. Audio is sent for transcription.
See the captions in motion
How it works
Add the narration
Add the narrated video or voiceover, choose its spoken language and sign in to generate captions. The narration audio goes to transcription, and the file itself stays on your device.
Decide who leads each scene
Correct the transcript first. Then decide scene by scene whether the words or the footage should lead.
Watch it with the sound off
Download the video and watch it once at phone size with the sound muted. The captions alone should carry the narration. Save the editable project too, in case a line needs fixing.

A faceless video can be a screen recording, an illustrated explanation or narration over footage. These need different caption placements. In each one the visuals carry information, so plan the captions around them.
LumaCaption styles the words from your audio or script. It does not research a topic, write a verified script or find suitable footage for you.
Build a clear first ten seconds
State the question or the result in words a new viewer understands without context. Show the thing you are talking about when you can. Then let the typography stress that one point.
For a tutorial, keep captions away from the cursor and the control being demonstrated. For a chart, avoid covering labels. For a quotation, verify it against the source.
Alternate intensity with rest
Open a section with full-frame type, then switch to calmer captions for the explanation. Watch the whole sequence and check that viewers have time to read and still look at the imagery.
Animated captions do not make a video original, accurate or eligible for monetisation. That depends on the material and on each platform’s rules.
- Text-led scenes
- Big Type styles let a short narrated point become the visual focus.
- Footage-led scenes
- Use a quieter style when a chart, product or screen demonstration needs room.
- Editable emphasis
- Choose the word in each cue that gets the accent, and leave the rest plain.
- Reusable outputs
- Save the transcript as TXT as well as the finished video, for reuse in a later edit.
Common questions
Do I need a talking-head recording?
No. Use a voiceover or another supported media file, and choose a backdrop if you need one. Then watch the pauses, where nothing new appears on screen.
Does this generate B-roll?
No. Bring your own footage, or use one of the editor’s backdrops.
Will full-frame text always perform better?
No. Full-frame text is a design choice and does not guarantee engagement. Put it over a screen recording and the viewer loses the demonstration.
