Word-by-Word Captions Like MrBeast and Hormozi (82 Styles)
October 5, 2026 · 7 min read · by the ClipCraft team
MrBeast captions are the ones you already know even if you've never heard them named: giant bold uppercase words, a hard black outline, and one word flashing yellow at the exact moment it's spoken. Alex Hormozi runs the other famous variant, where the active word gets a solid colored capsule behind it instead of a color change. Both are easy to recreate. ClipCraft's Video Captions tool ships 82 animated caption presets, including faithful takes on both looks, and every one of them is a starting point you can reshape into your own style. This post is about the styles themselves: why the word-by-word effect works, what each famous look is actually made of, and which of the 82 presets are worth your time.
Why one lit-up word keeps people watching
A static subtitle gives the viewer everything at once. They read the line in half a second, they're done, and their thumb starts drifting. A word-by-word highlight doles the sentence out on the speaker's rhythm instead. Your eye tracks the moving highlight the same way it tracks any motion, so attention stays parked on the frame until the line finishes, and then the next line starts and the loop repeats.
There's a second effect that matters just as much: most short-form video gets watched on mute. Animated captions carry the entire performance for those viewers, timing included. A well-timed highlight reads like the speaker's cadence. A badly timed one reads like a karaoke machine that had a stroke, which is why the timing half of this (covered near the end) needs as much care as the colors.
The anatomy of MrBeast captions, next to the Hormozi pill
Here are four of the presets on the same frame of the same video, frozen mid-sentence, so you can compare what each one is doing to the word "lights":

Beast Mode (top left) is the MrBeast recipe: Impact, forced uppercase, slightly oversized, with an outline that's a full 14 percent of the font size, which is what makes it readable over literally any footage. The active word turns pure yellow and scales up 1.18×, so it flashes and swells at the same time. Hormozi (top right) keeps the heavy uppercase text white and instead drops a rounded green pill behind whichever word is being spoken. The pill variant is calmer to read, which is why it took over business and podcast clips while the yellow flash stayed with stunt content.
Golden Karaoke (bottom left) is a different mechanism entirely. Nothing pops or changes color as a block; instead an amber fill sweeps left to right through the active word, and every word already sung stays filled. That mid-sweep frame shows it caught halfway through the word. Karaoke styles feel smoother than flash styles, and they're the natural pick for lyrics. Neon Sign (bottom right) shows what a glow-based style does: cyan text with a heavy blur bloom, and the active word shifts pink. Gorgeous on dark footage, nearly invisible on bright footage, so treat glow styles as a situational tool.
And then there's the sing-along classic. Bouncing Ball puts a little ball above the line that hops from word to word, and the preset deliberately forces the caption onto at least two lines so the ball always has somewhere to travel:

Touring the 82 presets without scrolling all of them
The gallery sits below the player, and every tile is a small canvas actually animating its style at around 30 frames per second, with the "spoken" word cycling, so you're auditioning real motion rather than a static thumbnail. Above the grid there's a search box that matches vibes ("horror," "retro," "cute") and a row of 15 tag chips. Each chip returns at least ten styles, so no filter dead-ends. The families worth knowing:
- The creator classics: Beast Mode, Hormozi, Red Alert (Hormozi's red sibling), Podcast Bar, and Pink Pop.
- Karaoke: ten styles built on the sweep, including Golden Karaoke and three bouncing-ball variants.
- Neon and glow: ten, from Neon Sign to Miami Neon and Laser. Dark-footage specialists.
- Horror and gothic: eleven horror styles on real display fonts (dripping blood, blackletter), plus a matching gothic set of ten.
- Gamer and techno: 8-Bit, Streamer, Boss Level, Cyber and friends, split between pixel and sci-fi fonts.
- Display and elegant: twelve quieter serif and script looks (Vogue, Big Serif) for content that shouldn't shout.

An honest note on picking: the loud styles are overrepresented in what people copy, and they aren't always right. A dripping-blood font on a cooking video is a bit, and bits age fast. The elegant pack and Clean Minimal exist because some content earns more trust with less costume.
Turning a preset into your brand style
The presets are starting points. Click any style, then open the Customize panel (the sideways tab next to your captions) and everything the preset is made of becomes a control: font, size, vertical position, text color, the highlight mode itself (color change, pill, karaoke sweep, underline, or the ball), highlight and outline and glow colors, word pop scale, and uppercase, bold, and background-box toggles. The moment you change anything, the style header reads "(edited)" and a Reset button appears, so experiments are free.

That screenshot is Beast Mode with exactly two changes: the highlight color swapped from yellow to violet, and the pop raised a touch. It stopped being MrBeast's look and started being a brand look, which is the actual point. Pick one preset, change the highlight color to your channel's accent color, and use that same combination on every video. Viewers learn it fast. If your edits land somewhere good, note your settings, because the customizer holds one working style at a time.
Two practical defaults worth knowing while you're in there. Captions sit at 78 percent down the frame out of the box, which is fine for widescreen but needs raising on vertical video (our Shorts captions guide covers the safe zones). And if a style vanishes over bright footage, the outline slider fixes it faster than changing styles.
Word timing: what's exact and what's estimated
The style is only half the effect. The other half is whether the highlight lands when the word is actually said, and the tool is upfront about having two tiers here. The ✨ Auto captions option transcribes your video with Whisper and returns per-word timestamps, so the highlight hits each word at the moment it's spoken. That costs 5 tokens per started minute of video (a free account includes 100 tokens a month, and token packs start at $0.99 if you need more).
Typed or pasted transcripts work differently. You can tap-sync each line's start for free, but the words inside a line get timed by their length, longer words get more time, which is a decent guess rather than a measurement. For a karaoke sweep or a bouncing ball, that gap is visible if you look for it. Flash and pill styles hide it better. Edit an AI-transcribed line to a different word count and it quietly falls back to the same length-weighted guess, so do your text cleanup before you obsess over sync.
If you want the full workflow walkthrough, the TikTok captions guide covers loading video, the transcript editor, and exporting step by step. And if your video is a song rather than a talking head, the Suno music video post shows the create-from-audio mode, where these same 82 styles run over a generated background.
My advice after building way too many of these: steal the mechanism, not the colors. Word-by-word highlighting is what holds viewers; the yellow is just MrBeast's paint. Pick the highlight mode that fits your content, recolor it to your brand, and stop touching it. A free account is enough to find your style today.
Find your caption style
82 animated presets, a full customizer, and word-level AI timing. Free to style and preview; export your first captioned video today.
Try the caption styles free