Titles and lower thirds in After Effects with AI
Titles and lower thirds in After Effects with AI: ask ChatGPT, Claude or any MCP client connected to Prism, and the name, role, chapter title or credit roll lands in your open comp as keyframed text and shape layers.
The lower third is the most templated object in After Effects: a pack of forty, one nearly right. Without the pack it is a sentence: who, what role, how it enters. Then the same style for every other name.
What you say, and what lands
| You type | What appears in your comp |
|---|---|
| "Lower third at 4 s: 'Maya Chen', 'Marine biologist' under it, bottom left. A 4 px orange bar wipes in, the name rises behind a mask, the role fades in after. Out at 9 s, reversed." | Two text layers and a bar shape layer, a mask reveal, position and opacity keyframes in and out |
| "Chapter title on black: 'PART TWO. THE ESTUARY', centred, tracking widens from 0 to 80 over 3 s while it fades up." | A solid, one text layer, tracking animator and opacity keyframes |
| "Location card, top left: 'Isle of Skye, Scotland', small caps, types on letter by letter with a thin rule under it." | A text layer with a write-on animator, a rule with a Trim Paths draw-on |
| "End credits from this list: roll bottom to top over 45 s, two columns, role left and name right." | Two text layers, one per column, position keyframes with a linear ease |
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
One style, many names
| You ask | What lands |
|---|---|
| "Here are the twelve interviewees and their timecodes; build a lower third for each in this comp." | Twelve pairs of layers at their times, one style |
| "Make each one its own comp so the editor can drop them in." | One comp per name, named after the person |
| "Change 'Marine biologist' to 'Marine ecologist' wherever it appears." | Every matching text layer in the project rewritten |
| "Swap every title to Inter-SemiBold." | Every title layer switched, project-wide |
Timing to the edit
| Timing comes from | You ask | What lands |
|---|---|---|
| markers you placed | "Put a lower third on each named marker." | one title per marker, in on the marker |
| the interview audio | "Transcribe it, and place each person's lower third the first time they speak." | transcript markers, a title at each voice's first line |
| the music | "Land each chapter title on a downbeat." | beat markers on the timeline, titles snapped to them |
Transcription can separate speakers but does not know their names; you say which voice is which.
What it needs
| After Effects | 2024, 2025 or 2026, macOS or Windows |
| AI client | any MCP client: ChatGPT, Claude, Claude Code, Cursor, Codex and others |
| Plan | a Free account is enough for the titles; transcription and beat detection are Prism AI capabilities, paid from your balance |
What it does not do
It does not ship a pack of lower thirds, read names from a script or spreadsheet file, or recognise who is on screen. You paste the list and say who speaks when.
Common questions
Related
SaaS product
SaaS product animation in After Effects with AI: export the UI from Figma as SVG, import it with live text, and describe the cursor, panels and callouts that move.
Presentations
PowerPoint to After Effects with AI: paste a slide's content or its SVG export and it is built in your comp as editable text and shape layers you can animate.