Plenty of people watch Reels with the sound off: on a bus, in a waiting room, in bed next to someone who's asleep. If your Reel only makes sense with audio, you lose those viewers in the first few seconds. On-screen text and subtitles fix that, and they also help viewers who are deaf or hard of hearing, or who are watching in a second language.
Two kinds of on-screen text
Most good Reels use two layers of text, and it helps to plan them separately.
- Headline text tells people what the Reel is about. It appears in the first second or two, uses a few large words, and often stays up for the first part of the video. For example: "3 ways to fold a shirt for travel."
- Subtitles follow what's being said, line by line, in sync with the audio.
There's also a technical split worth knowing:
- Burned-in (open) captions are part of the video itself. Everyone sees them, and you control the font, size, and position.
- Closed captions are generated or added separately, and viewers choose whether to turn them on. Instagram can generate these automatically for Reels with speech.
Burned-in subtitles are the safer default, because they show up for everyone and look the way you designed them. If you burn them in, viewers who have Instagram's own captions switched on may see the text twice. Instagram has offered a per-Reel setting for auto-generated captions in the advanced settings before you share; if your app shows it and your burned-in subtitles are accurate, you can turn the automatic ones off for that Reel.
Auto-captions and how to fix them
You rarely need to type subtitles by hand anymore.
- In Instagram's editor, the captions option transcribes your speech and puts it on screen. It has lived in the sticker menu and, in some versions, in the editor's toolbar. Tap the captions to edit words, change the style, and drag them into position.
- Dedicated editing apps have their own auto-caption tools, usually with more styling and timing control. If you edit outside Instagram, the guide on editing Reels on your phone covers the full workflow.
Automatic transcription is a first draft. These are the parts it most often gets wrong:
- names of people, places, businesses, and products,
- numbers, prices, and measurements, such as "fifteen" heard as "fifty,"
- words that sound alike, like "their" and "there,"
- speech over music, background noise, or with a strong accent,
- slang and industry terms.
Check them in two passes. First, play the Reel with sound on and pause on each line to compare it with what was said. Then watch it once fully muted and ask whether it still makes sense. A small business that posts the wrong price in a subtitle has a bigger problem than a typo.
Make captions easy to read
Readable subtitles come from a few consistent choices.
- Show one phrase at a time. One or two short lines per screen is plenty. Break lines at natural pauses, not in the middle of a phrase like "the / best."
- Leave text up long enough. If you can't comfortably read a line twice in the time it's on screen, shorten it or let it stay longer. Fast word-by-word pop-ups look energetic but can be tiring to follow; phrase-by-phrase is calmer.
- Use a bold, simple font. Clean sans-serif fonts hold up at small sizes. Avoid thin, script, and heavily stylized fonts for subtitles.
- Go big enough for a phone. Judge size on your phone at normal viewing distance, not in a zoomed-in editor preview.
- Add contrast. White text with a dark outline or shadow, or text on a solid box, stays readable on any background. Colored text on a busy scene usually doesn't.
- Use sentence case. It reads faster than all caps for anything longer than a few words. Save capitals for one- or two-word emphasis.
- Highlight sparingly. Coloring one key word per line can help. Highlighting every other word just adds noise.
Place text inside the safe zone
Instagram's interface covers parts of your video, so text placement matters as much as text style. On a 1080 × 1920 canvas, keep text clear of roughly:
- the top 250 px, where the Reels header and camera icon sit,
- the bottom 400 px, where your username, caption, and audio name appear,
- the right 150 px, where the like, comment, and share buttons are.
Within that space:
- Subtitles work well in the lower-middle area, roughly between 1,250 and 1,500 px from the top. That keeps them above the caption area without covering faces.
- Headline text fits in the upper-middle, below the top margin. If you might use that frame as your cover, keep it inside the center 3:4 area so it survives the profile grid crop.
- Don't stack layers. If a headline and subtitles are on screen at the same time, give them separate zones so they don't collide.
Plan for this when you film. Framing your face in the upper half of the shot leaves room for subtitles below your chin; filming Reels with your phone has more on framing. The Reels size and specs guide explains the safe zones in more detail.
Accessibility beyond subtitles
Subtitles are the biggest step, but a few more habits make your Reels work for more people.
- Caption sounds that matter. If the doorbell, a laugh, or the music cutting out is part of the story, add it in brackets, like [doorbell rings].
- Say who's speaking. With more than one person on screen, add a name label or a dash before each new speaker. Don't rely on text color alone to tell speakers apart.
- Don't let "this" do all the work. If you say "this one is better" while pointing, add text that names the item.
- Avoid rapid flashing. Strobe-style text effects and quick flashes can be uncomfortable or unsafe for some viewers.
- Write a useful post caption. Screen readers read the post caption, so describe what the Reel shows in plain words instead of only using hashtags or emoji. Don't swap key words for emoji; screen readers read emoji names out loud.
Quick checklist
- Headline text appears in the first second and names the topic.
- Subtitles are proofread, especially names, prices, and numbers.
- One or two short lines per screen, each up long enough to read twice.
- Bold font, strong contrast, sentence case.
- Text sits inside the safe zone and off faces and products.
- The Reel makes sense when watched fully muted.
- The post caption describes the video for people using screen readers.