Captions
Pick from 57 caption presets, override any style detail, or turn captions off.
Every clip renders with animated captions driven by its transcript. You style them in the editor; new clips start with the caption settings of the Brand that processed the Video.
Presets
OpenClip ships 57 caption presets. Each one is a complete look: font, colors, stroke, shadow, position, and animation. A few examples:
| Preset | Style |
|---|---|
| Clean | The default: bold white Inter with a crisp black outline and a soft shadow, in the lower third |
| Minimal | Lowercase with a hairline outline and no accent |
| Plain | White text on soft dark bars, like broadcast subtitles |
| Quiet | Black text on white bars, one bar per line |
| Dan | Condensed caps with a thick outline; the second line runs yellow |
| Mozi | Heavy caps; the spoken word flashes green |
| Karaoke | Words wait in grey and light up yellow as they are spoken |
| Neon | The spoken word glows electric cyan |
Pick a preset in the Caption presets tab of the editor's library. The preview updates immediately.
Motion presets
Seventeen presets are built on the motion engine and sit at the top of the picker: every word gets its own entrance timed to the voice, and the spoken word gets a signature treatment.
| Preset | Motion | Type |
|---|---|---|
| Chase | A lime box glides from word to word, stretching between them | Plus Jakarta Sans ExtraBold caps |
| Sweep | True karaoke: each word fills left to right as it is spoken and stays lit | Montserrat Black caps |
| Haze | Serif italic words materialise out of a soft blur | Instrument Serif italic |
| Stamp | Words slam in on the beat with a small camera shake | Barlow Condensed Black caps |
| Jelly | Every letter springs up, overshoots and squashes as it lands | Fredoka Bold lowercase |
| Retro | Cream 3D-extruded diner lettering that pops in word by word | Shrikhand |
| Chrome | Polished chrome caps with a glint across the spoken word | Big Shoulders Display ExtraBold caps |
| Lens | Rack focus: the line sits soft and focus pulls onto each word | Manrope ExtraBold |
| Swipe | A highlighter swipes under each spoken word and stays | Schibsted Grotesk ExtraBold |
| Glass | A frosted glass plate; the spoken word brightens and lifts | Inter Tight SemiBold |
| Stencil | White outline caps that fill up left to right as each word is spoken | Anton caps |
| Solo | One huge word at a time, springing in on every beat | Unbounded ExtraBold lowercase |
| Decal | Candy-coloured sticker caps with a white die-cut border, each at its own angle | Lilita One caps |
| Rail | A bar glides under each word as it is spoken | Hanken Grotesk ExtraBold |
| Glitch | Red and cyan channels tear in and snap together on the beat | Chakra Petch Bold caps |
| Tube | Neon tubing that stutters on word by word | Tilt Neon lowercase |
| Dusk | A sunset-gradient serif; the spoken word glows up to white | DM Serif Display italic |
Motion presets keep their own choreography, so the animation mode overrides below do not apply to them. Font, colors, stroke, shadow and position overrides still do. The highlight color drives the accent: the Chase box, the Rail bar, the Sweep and Stencil fill, the Swipe marker and the Stamp flash.
Style overrides
You are not locked into a preset. On top of any preset you can override individual properties, including:
| Property | Options |
|---|---|
| Position | Bottom center, lower third, center, upper third, top center |
| Animation mode | Per word or per segment |
| Word animation | Static, pop, kinetic, burst, fade, bounce, reveal, slide, shake, glow-pulse |
| Segment animation | Static, typewriter, kinetic, burst |
| Typography | Font size, weight, line height, text transform |
| Colors | Text, active word, inactive word |
| Outline and depth | Stroke width and color, shadow blur and color |
Overrides are stored per clip, so each Viral Moment can have its own look.
A bigger font shows fewer words at a time. When a caption no longer fits its preset's lines, it plays in short parts, each on screen from its first spoken word, so no word is skipped. A single word too wide for the frame, like a long name or a web address, shrinks until it fits.
Emphasis words
The AI picks the words that carry each clip, like names, numbers and punchlines, and captions color them with the preset's highlight color, in the editor preview and in every export. Presets without a highlight color of their own, like Clean, show emphasis words in yellow. Change the highlight color to recolor them.
Emoji
The AI also picks an emoji for the key words in each clip. Turn them on with the switch in the editor's Emoji tab, which lists the emoji picked for that clip, each beside the caption line it shows on with the word that set it off (click one to play from there); there you also set the emoji style, animation, position and size, and the style and animation menus show each option on one of your clip's emoji. On the timeline, each caption page shows the emoji it carries. Your emoji settings stay as they are when you switch caption styles.
An emoji set above the text shows below it when there is no room above, as with captions at the top of the frame, and the other way round.
Bangers
Bangers set a clip's punchiest lines big and stacked behind the speaker: the lead-in words small on top, the punch word huge underneath, with the speaker's head in front of it. Turn them on with the switch in the editor's Bangers tab, and pick a style:
| Style | Look |
|---|---|
| Captions (default) | Your caption style, blown up: its font, casing and outline, the lead-in in its text color and the punch in its highlight color |
| Stack | Condensed caps in the caption preset's highlight color, a small bold lead-in |
| Serif | An italic serif lead-in over heavy white caps |
| Outline | Outlined caps stacked tight, the last line filled with the highlight color |
| Boxed | Bold caps on color slabs |
| Flank | The words zigzag down either side of the speaker, behind them, the punch in the highlight color |
| Crown | The punch huge across the top, the speaker's head in front of its foot, the lead-in beside the head below it |
OpenClip's AI picks the lines from what the speaker says (the hook, then the reveals, numbers and strong words a viewer would screenshot) and times them to the speech: the lead-in words appear as they are said and the punch word slams in on its beat. A clip gets the hook and then about one banger every seven seconds. The switch shows how many the clip got, and the timeline marks each one.
The Behind slider under the styles sets how far behind the speaker the lines go: at 0% the punch word sits on top of the head, at 100% the head reaches the top of it. It is also the most of any line the speaker may cover: a line they would cover more of (common in close-ups, where the head fills the top of the frame) is drawn in front of them instead, so every line stays readable. In shots with nobody on screen, the line shows in front. The regular captions pause while a banger's line is said.
Flank and Crown swap the Behind slider for Pop-out: while a banger is up, the background blurs and darkens and the speaker shrinks a little to make room beside them, then everything eases back. Off keeps the shot as it is, Pop shrinks the speaker a little and Pop+ a little more. Flank starts on Pop and Crown on Pop+. When the speaker fills the frame, like a face in a tight close-up, there is no room beside them: those bangers stack like the other styles, so every line stays readable.
Bangers need the speaker cut out of the clip. The first time you switch Bangers on for a clip, the card offers Cut out the speaker. It is free and runs right in your browser, on your computer's graphics chip: a glowing outline over the preview traces the speaker as it goes, and a one-minute clip takes about 15 seconds on a recent laptop (longer on older computers). Keep the tab open until it finishes; the preview shows the bangers as soon as it is done. It needs a browser with WebGPU, like Chrome, Edge or Safari on a computer. You cut a clip out once: switching bangers on and off or changing their style keeps it. If the clip is reframed again, its new footage needs a new cut-out. Your banger settings stay as they are when you switch caption styles.
AI agents (through MCP) and the API can switch bangers on, but the cut-out only happens in the web editor: open the clip there once and press Cut out the speaker. Until then the clip renders with its plain captions. Agents get each clip's editor link as edit_url from list_clips.
Turning captions off
Captions can be toggled off entirely for a clip. The clip then exports without captions. Your trims and silence removal still apply.
Fixing caption text
Captions render whatever the transcript says. If a word is wrong, correct it in the editor's Words tab; see the editor. Timings are untouched, so the animation stays in sync.
Free plan limits
On the Free plan, two presets are available: Clean (the default) and Dan. The others show a lock in the editor, and clicking one offers the upgrade; a clip on another preset moves to Clean when you open it. Free plan exports also always carry the OpenClip watermark. Upgrade to remove the watermark and unlock all 57 presets; see credits and plans.
Captions in other tools
The same preset system powers Quick Captions and your Brand defaults, so a look you like can become the starting point for every future clip.