Caption Styles: 17 pro caption looks with presets — effect skill for captions & subtitles
For AI agents: find the right skill in one call —
curl -s "https://varg.ai/how-to/api/search?q=<what the person wants>"returns the matching how-to page and the best skills with install commands, as markdown (Russian works too). Full index: https://varg.ai/how-to/llms.txt. Every page has a markdown twin: add.mdto its path. Before generating anything, show the person the options on the how-to page. From a terminal:curl -fsSL https://varg.ai/install | shinstalls the vargai CLI (https://varg.ai/how-to/cli.md):vargai searchandvargai install <slug>with no key, every varg model aftervargai login.
Karaoke, word pop, pill, glass, neon and 12 more caption looks as exact presets: sizes, safe zones, sync and contrast from subtitle pros.
- Type: Effect · Captions & subtitles
- Slug:
caption-styles - Status: live
- Publisher: varg (varg, verified)
- License: MIT
- Source: https://github.com/vargHQ/skills/tree/main/caption-styles
- Works in: Claude Code, Codex, Cursor, Gemini CLI, OpenClaw, varg Studio
- Requires: Node 22 + ffmpeg (HyperFrames) or a Remotion 4 project
- Cost: Free: renders locally with HyperFrames, Remotion or ffmpeg. Generated footage, voice or music is billed in varg credits with an estimate first.
- Runtime: Minutes per template, mostly render time
What it does
Seventeen caption looks with presets you can apply as numbers, not adjectives. Pick how the words are chunked first (a phrase, a group of words or one word at a time), then the look. Every look ships with a plain version that is accessible, since styled word-tracking captions trade some comprehension for attention.
17 templates, 52 presets. Every rule in the skill cites its source; every preset is exact values (durations, easing, sizes, colors, fonts) in a JSON file the agent loads.
| Template | Presets |
|---|---|
| Active word colour captions: White outlined text, a few words at a time, where only the colour of the spoken word changes: no scale, no box | Blue outline, Yellow outline, Calm phrase (accessible) |
| Active word pop captions: White outlined words where the word being spoken turns blue and pops a little larger, then settles back | Blue pop, Subtle pop, Spring pop |
| Bouncing single-word pill: One uppercase word at a time bounces up inside a coloured pill (purple by default) near the centre of the frame | Purple spring, Lime spring, Soft rise (no overshoot) |
| Frosted panel captions: White text on a blurred, frosted-glass panel; inactive words are dimmed and the spoken word is full white | Clear glass with dim, Thick glass for bright footage, Light glass |
| Gradient-fill word captions: The spoken word fills with a two-colour gradient (blue to pink by default) inside a dark rounded box while the other words stay white | Blue to pink on dark box, Sunset, Sweep fill |
| Grow-in highlight box captions: White words inside a dark rounded box; the spoken word turns yellow and grows into place while the box adjusts its width | Yellow grow, Green grow, Reveal grow |
| Karaoke fill captions: Each word fills with yellow as it is sung or spoken, with a short bar under the current word | Classic yellow sweep, Two-line preview, Speech (instant fill) |
| Line-hugging rounded box captions: Bold dark text on a white box whose rounded corners follow each line's own width, the native look of in-app text on vertical platforms | White box classic, Black box inverse, Navy pill rise, Brand tint |
| Neon tube captions: Words glow like neon tubes (hot pink by default) inside a dark box with a glowing border | Hot pink, Cyan ice, Pulse only (no flicker) |
| Outlined spring-word captions: White words with a thick black outline; the spoken word turns cyan and springs up a few pixels, then settles | Cyan spring, Gaming condensed, Clean outline |
| Serif italic emphasis captions: Large serif text with a soft shadow and no box; the spoken word flips to italic | Instrument italic, Playfair warm, Fraunces soft |
| Shadow-only minimal captions: White text with a soft shadow and no box; the spoken word is just a little brighter than the rest | Soft shadow, Cinema lower, Bright-footage guard |
| Sliding pill word tracker: A coloured pill slides from word to word behind white outlined text, marking the word being spoken while the text stays still | Electric blue, Brand pill, Soft pill |
| Tilted card word pops: One uppercase word at a time springs in on a slightly rotated sticker-like card, alternating white and lime | White / lime alternating, Sticker with hard shadow, Calm tilt |
| Translucent box subtitles: Plain white sentence-case text on a semi-transparent dark box, one phrase or sentence at a time, cut on and cut off with no motion | Broadcast 16:9, Vertical 9:16, Soft off-black |
| White card captions: Dark text on a clean white card with a soft shadow; the spoken word turns a strong blue | Corporate blue, Warm paper, Vertical card |
| Word tile captions: Every word sits on its own dark rounded tile; the spoken word's tile turns orange | Orange active, Stagger in, Mono tiles |
Rules every template follows
- Choose chunking first: phrase (5–12 words), group (2–5), single word only for hooks/punchlines/≤20 s clips (word-for-word: ~90% gaze on text, lower comprehension — Romero-Fresco; Meryl Evans).
- Active word switches 1–2 frames before audio onset (late noticed from ~45 ms, early unnoticed to ~125 ms — ITU-R BT.1359). Karaoke: line leads 10–20 f (≤1.5 s), fill follows voice (\kf).
- Per-word motion done within ~6 f at 30 fps; scale via transforms, never font-size.
- 9:16 safe box: top 288, bottom 672, left 120, right 192 px on 1080x1920; max width 696 px, block centre ~60% height.
- Contrast ≥4.5:1 against what is behind; black box ≥60% opacity on white footage.
- Netflix/BBC/DCMP line breaks; 42 cpl 16:9; 2 lines (3 in 9:16); 160–180 wpm; 0.83–7 s per caption; 2-frame gaps.
Built from: BBC Subtitle Guidelines; Netflix Timed Text Style Guide; Meryl K. Evans (@merylkevans); DCMP Captioning Key; ITU-R BT.1359; Romero-Fresco; WCAG; X: @wayofzorro, @finnardley, @mws, @yt_igm, @tigerjvideo.
Renders locally with HyperFrames or Remotion for 0 credits; generated footage, voice or music runs through varg with a cost estimate first.
Use when
- How to add word-by-word captions to a video
- How to make TikTok-style highlighted captions
- Karaoke subtitles that fill as you speak
- Best font size for captions on Reels and Shorts
- Caption safe zone for TikTok, Reels and Shorts
- How to make captions readable over any background
Not for
- Transcribe this interview to an SRT file (use whisperx or elevenlabs-speech-to-text)
- Translate my subtitles into Spanish (use gemini-srt-translator)
- Add a lower third with the guest's name (use title-design)
Install
- Claude Code:
claude plugin marketplace add vargHQ/skills && claude plugin install varg@varg - Codex:
npx skills add vargHQ/skills --skill caption-styles -a codex -y - Cursor:
npx skills add vargHQ/skills --skill caption-styles -a cursor -y - OpenClaw:
npx clawhub@latest install caption-styles - Gemini CLI:
npx skills add vargHQ/skills --skill caption-styles -a gemini-cli -y - Any agent, with the vargai CLI:
vargai install caption-styles
Example output
https://s3.varg.ai/files/72c55813-c0fa-4cf2-9f64-ec4f7ed0e7ae/file_4d3d0dbdc222.mp4 — example (credit: varg)
More captions & subtitles skills
- 3D Camera Captions: Captions that live in 3D around a talking head: the camera flies through the words, heroes pass behind the speaker.
- Remotion Captions: Transcribe speech and burn in animated captions with Remotion, or import an existing SRT file.
- ElevenLabs Speech to Text: Transcribe audio and video with word timestamps, speaker labels and keyterms, ready for subtitles.
- ElevenLabs Speech to Text (inference.sh): Transcribe audio with ElevenLabs Scribe, speaker diarization and forced alignment, via inference.sh.
- Speech to Text: Transcribe audio with ElevenLabs Scribe v2 or Whisper V3 Large, with timestamps and speaker labels.
- Pretext Text Metrics: Measure and fit text without the DOM using pretext: auto-fit sizes, know line breaks before render, canvas/WebGL text.