Burned-In Captions for Shorts, and the Settings That Matter

Do captions actually matter on a 30-second video?

They matter more on short videos than anywhere else. Think about how people watch Shorts and Reels: on a bus, in a queue, in bed next to someone who's asleep. Lots of them have the sound off. If your words only live in the audio, those viewers get moving pictures and no story, and they swipe away.

Captions also help people who do have the sound on. Reading along gives the eye something to follow during the slower moments, and short-form lives or dies on those moments. A viewer who's reading is a viewer who's still there.

So it isn't really "should I add subtitles?" It's "how do I add them without spending an hour per clip?" That's the part most people get stuck on.

Burned-in captions or the platform's auto-captions?

This is the first decision, and it's the one people usually can't judge on their own.

Platform auto-captions are the ones TikTok, Instagram or YouTube generate after you upload. They cost nothing and need no editing. The catch is that you don't control them. The viewer may have to switch them on, the styling is generic, and the transcription can mishear names or niche terms.

Burned-in captions are part of the video file itself. Everyone sees them, on every platform, styled the way you chose. If you post the same video to several platforms, they look identical everywhere. The trade-off is that you can't change them after export without re-rendering.

For faceless channels, burned-in wins almost every time. The text is the visual hook. You want it on screen from the first frame, not hidden behind a toggle.

The four choices that actually change how captions read

This is the core of it. Fonts get all the attention, but these four settings decide whether your captions help or get in the way.

Chunk size. Show a few words at a time, not whole sentences. A full sentence at the bottom of a vertical screen is a paragraph to someone scrolling. Short chunks keep the eye moving with the voice. Up to four words per chunk is a comfortable rhythm for most narration.

Position. Bottom is the habit, but on vertical video the bottom and right edge are crowded with the username, the caption text and the like and share buttons. Check one of your videos on your actual phone, inside the app, before you post a batch. If the platform's interface covers your words, move them up.

Colour and size. White text with strong contrast works for nearly every niche. Size it so you can read it at arm's length without squinting. Save the bright colours for channels where the brand calls for it, like comedy or kids' content.

Word highlight. That's the effect where each word lights up as it's spoken. It suits fast, punchy niches like motivation, facts or quizzes. For sleep stories, history or anything calm it can feel jumpy. Plain chunks are often the better choice there, and nobody will think anything is missing.

How this works when the video is generated for you

If you make faceless videos with MaqtAi, captions are built in. They come from the same script the AI voice reads, so the words and timing match the narration with no transcription step in between.

By default, captions appear in chunks of up to four words. To change how they look, there's one thing to know first. The video wizard starts in Simple mode, which picks the voice, sound and captions automatically and hides those steps. Switch the toggle at the top of the wizard to Advanced, then open the Captions step (its page heading is "Caption Style"). There you can:

• turn captions on or off
• pick the font and size
• set the position (bottom by default)
• change the text colour
• switch on "Highlight words as they're spoken" (off by default)

If you're already on the Review step, "Change these" switches you to Advanced. Then open Captions from the step bar at the top.

The caption language isn't set on that step. It follows the voiceover language picker on the Topic step, which controls the script, narration and captions together. Pick the language there before the script is written.

What this won't do (read this before you commit)

Now the honest part. This isn't a tool for captioning footage you filmed yourself. If you shoot talking-head clips on your phone, your editing app's auto-caption feature or the platform's own captions will serve you better. MaqtAi captions the faceless videos it creates.

A few other limits:

• No separate caption language. Captions always match the narration language. You can't have English voice with Spanish subtitles in one render. If you want a second language, the Language versions panel on a finished video makes a new voiceover and captions in that language, with the visuals unchanged.
• Word highlight is refused for Arabic and other right-to-left scripts. Captions still render normally there and still follow the voice. They just don't do the per-word highlight.
• Captions show exactly what the script says. If a name or term is spelled wrong in the script, it's spelled wrong on screen. Read your lines on the Scenes step before you launch.
• Captions don't make a video good. They remove a friction. A weak hook with perfect captions is still a weak hook.

Will people think it looks cheap or obviously AI?

This is the worry nobody types out. The honest answer: burned-in, chunked captions are standard on short-form now, whether a person or a tool made them. Viewers judge whether the first line grabs them and whether the story pays off. They don't care how the captions were timed.

What does look cheap is the same caption style, the same voice and the same structure on every video in a feed. Change it up across series. Use a different colour or font per niche and keep the highlight for the videos that need the energy.

What waiting actually costs

Every week you spend deciding on a caption workflow is a week you don't post, and posting is how you learn anything. The algorithm learns nothing about a channel with no videos. Meanwhile someone in your niche posted today, and their captions were fine. Not perfect, just fine.

The smallest useful step is one video. Pick a topic, switch the wizard to Advanced, set your caption style once, and watch the result on your own phone. Nothing is generated or charged until you press Launch. Most new accounts that verify their email get free credits with no card needed (details on pricing; free-tier renders carry a watermark). You'll know in an afternoon whether this style fits your channel.

Frequently asked questions

Can I add subtitles to a video I already filmed? Not with MaqtAi. It captions the faceless videos it generates. For your own footage, use your editing app's auto-captions or the platform's built-in ones.

Where should captions sit on a vertical video? Away from the bottom and right edge, where the app's buttons and text sit. Bottom is the default, so check one video in the actual app and move the captions up if anything covers them.

Do I need the word-by-word highlight effect? No. It's optional and off by default. It suits fast, high-energy niches, and plain chunks of a few words often read better for calm or story-driven content.

Can captions be in a different language from the voice? No, captions always match the narration language set on the Topic step. To reach another language, make a language version of the finished video, which changes the voiceover and captions together.