Skip to content
Prism
Use cases

Word by word subtitles in After Effects with AI

Yes, you can make word by word subtitles in After Effects with AI: ChatGPT, Claude or any MCP client connected to Prism transcribes the clip and builds timed text layers in your open composition, the spoken word highlighted.

The Hormozi look, the TikTok pop, the Shorts karaoke line: every page that ranks for them sends you out of After Effects. There are no presets here. Describe the style; the AI builds it as text layers you can change.

What you say, and what lands

You typeWhat appears in your comp
"Transcribe this clip, then make Hormozi-style captions: Anton, all caps, white with a thick black stroke, three words at a time centred low in frame, the spoken word pops to yellow."One text layer per three-word group, in and out on the words, a text animator that turns the current word yellow on its timestamp
"One word at a time, big, centre frame, each word scales in from 80%."One text layer per word, timed to its start and end, a scale pop keyframed on each
"Same captions, but Montserrat Black, and highlight with a rounded box behind the word instead of a colour."The layers restyled in place, a shape-layer box moving with the highlighted word

Test it on your own project. A Free account is enough to start: Sign up, then connect your AI and ask for this.

How to do it

1
Open the comp with the clip
The talking-head video or its audio must be a layer in the open composition.
2
Ask for the transcript
Select the layer and say "transcribe this". A [Transcript] guide layer appears above it with one marker per word.
3
Describe the style in one sentence
Font, case, stroke, words per group, position, and what the spoken word does. That sentence is the preset.
4
Review it in the timeline
One text layer per group, timed to the words. Nudge an in point or fix a misheard word.
5
Restyle by asking
"Yellow to green", "two words instead of three": the same layers are rewritten, and the transcript is not run again.

The style, named

ElementThe usual choiceHow to ask
FontAnton, Montserrat Black or Bebas Neueby PostScript name, no spaces, for example Montserrat-Black
Caseall caps"all caps"
Strokethick black, sometimes yellow"thick black stroke"
Words on screen1 to 3, at most 4 to 6 across two lines"three words at a time"
Highlightthe spoken word in yellow or green, or a box behind it"the spoken word pops to yellow"

One transcription, many styles

Transcription is a Prism AI capability, paid from your balance. Every restyle afterwards reuses the word markers and costs MCP actions only.

What it needs

After Effects2024, 2025 or 2026, macOS or Windows
AI clientany MCP client: ChatGPT, Claude, Claude Code, Cursor, Codex and others
PlanPro with a Prism AI balance for the transcription; the caption layers use MCP actions

What it does not do

It does not ship a caption preset pack, and it does not import an SRT or copy a style from a reference video. The look is a description; a second look is a second sentence.

Common questions