How do I add text to a video with AI?

Upload the video to Valmera and describe the text, where it goes and when: "add a lower third reading 'Sara Chen, Product Lead' for the first four seconds". The agent places it in clear space, animates it, renders a preview and checks the frames. Revise in plain English, then export with the text burned in.

What is kinetic typography, and can Valmera generate it?

Kinetic typography is text that moves in time with speech or music. Valmera generates it from your transcript: ask "animate my key phrases as I say them" and each phrase appears on the instant it is spoken, placed in the empty space beside the speaker, alternating sides, with emphasis words in an accent color. Bottom captions are hidden under it so words never print twice.

Can I animate text with keyframes?

Yes. Any text item can carry its own position, scale, rotation and opacity curves with easing, such as "slide in from the left, hold, then drift up and fade out". Keyframed motion replaces the named entrance and exit, and it travels with the text if you move its timing.

Which fonts and templates are available?

Seven templates (title, subtitle line, lower third, callout, big number, quote and chapter) and bundled fonts including Inter Display, Anton, Bebas Neue, Archivo Black, Poppins Black, Syne, Playfair Display, Instrument Serif, DM Serif Display, Montserrat and Plus Jakarta Sans. Custom font uploads are not supported.

Why did my captions disappear under a title?

Designed text hides the spoken-word captions while it is on screen, so two layers of words never stack. Remove or move the text and the captions return. To keep both, ask for the captions to stay visible and place the text in a separate area.

Can text follow a moving object?

Not by tracking. Text is placed in the frame and can be animated along a path you describe, but it does not lock onto a moving object. To put words behind a person as they move, use text behind the subject.

Creating an account and uploading are free. Editing requires a subscription. Exports are full-resolution H.264 MP4s and end with a brief Valmera end card.

Add Text to Video: Kinetic Typography, Titles and Text Animations

To add text to a video with AI, upload it to Valmera and describe what you want: “a lower third with my name at the start” or “animate my key phrases on screen as I say them.” The agent places designed titles, lower thirds and callouts, or choreographs kinetic typography that appears on the instant each phrase is spoken, in the clear space beside the speaker.

Every element is editable text with its own timing and motion, from named entrances to keyframed paths, so revisions are one sentence each.

Put Words on Screen by Describing Them

Upload, describe the text, review the preview. Creating an account and uploading are free; editing requires a subscription.

Create an account →

See pricing →

How to Add Text and Kinetic Typography in Valmera

1
Upload and cut first
MP4, MOV, M4V, MKV or WebM, up to 14 GB and 3 hours. Make your cuts before adding text so every title lands on the final timing.
2
Pick the treatment
A designed title or lower third for names and labels, kinetic typography for spoken phrases, a title card to break chapters.
3
Give the exact words and moment
Quote the text and say when: “from 0:32 to 0:35” or “when I say ‘pricing’.” Add font, color, position or motion if you have a preference.
4
Review the frames, then export
The agent renders a preview and checks the text against faces, captions and the frame edge. Adjust in plain English and export a full-resolution MP4.

PROMPTS TO PASTE

Add a lower third reading “Sara Chen — Product Lead” for the first four seconds.

Animate my key phrases on screen as I say them, beside my head.

Build the line “less editing, more making” word by word as I say it, with “more making” in serif italic.

Put a big “10x” on screen from 0:32 to 0:35 with a pop entrance.

Slide the title in from the left, hold it, then drift it up and fade it out.

Planning several text items? Download the on-screen text brief to note wording, timing and style for each.

Text and Motion Graphics You Can Ask For

Treatment	What it does	Ask for
Titles and templates	Title, subtitle line, lower third, callout, big number, quote and chapter marker	“A lower third with my name at the start.”
Kinetic typography	Spoken phrases appear on the instant they are said, in the empty space beside the speaker, alternating sides	“Animate my key phrases as I say them.”
Typography scenes	One phrase in mixed fonts, built word by word on speech cues, up to six lines	“Build ‘less editing, more making’ as I say it.”
Keyframed motion	Position, scale, rotation and opacity curves with easing on any text item	“Drift up slowly and fade out.”
Title cards	A full-screen card spliced into the video, then back to the footage	“A black title card saying ‘Chapter Two’ at 1:10.”
Vector graphics	Panels, underlines, arrows, rings and progress bars that share the text's motion	“Circle the Export button with a ring.”
Freeze frame with text	The picture holds, softens and darkens behind big centered words	“Freeze on 2:05 with ‘Day One’ over it.”

How Kinetic Typography Works in Valmera

Kinetic typography is text that moves in time with speech. Valmera builds it from the word-level transcript, so there is no manual keyframing: ask for it over a range or the whole video and each phrase appears at the instant it is spoken, placed in measured empty space around the speaker (above, below or to the sides), and holds until the next phrase replaces it. Emphasis words get the accent color, a size bump and a pop.

The default motion is one restrained settle across the whole passage, with a controlled overshoot saved for emphasis, so it reads as choreography rather than a roulette of effects. Ask for “still” when cutting carries the energy. For a single designed line, a typography scene mixes fonts (a bold sans with a serif-italic accent, say) and builds it word by word on the spoken cues; earlier words never jump as later ones arrive. Each phrase is an ordinary text item you can restyle, retime or remove.

Settings That Matter

Setting	Default	How to change it
Position	The template's default, in clear space	“Top left,” “beside my head,” or an exact spot.
Size	Set by the template	“Bigger,” or a scale from 0.4x to 3x.
Font	Set by the template	Inter Display, Anton, Bebas Neue, Archivo Black, Poppins Black, Syne, Playfair Display, Instrument Serif, DM Serif Display, Montserrat or Plus Jakarta Sans.
Color and backing	Template colors	Text and accent colors, a backing panel, outline and shadow.
Entrance and exit	Set by the template	Fade, pop, slide up, blur in, whip, rise, drop, elastic, bounce, swing, zoom blur, typewriter (entrance only) or none.
Timing	Where you place it in the edited video	“From 0:32 to 0:35,” or “when I say ‘pricing’.”
Captions underneath	Hidden while the text is on screen	“Keep the captions visible under the label.”

Common Mistakes

A different style for every title. One type system per video. Mixed templates with unrelated motion read like a slide deck.
Text over a face. Put words in the clear space (sky, wall, the side of the frame). The agent checks this; say so if a frame still collides.
Too little reading time. Watch when the last word appears, not when the animation starts. Shorten the line or extend the window.
Kinetic text on every minute of a long video. It shines in hooks and short-form; on a 20-minute lesson, plain captions are easier to follow.

Limits

Fonts come from the bundled set; custom font uploads are not supported. Each text element holds up to 200 characters. Text does not track moving objects, though it can follow any path you describe. Text is burned into the export, so changes after download mean exporting again from the project.

Related Tools

For words behind a person, see text behind the subject. For the spoken words themselves, use captions or animated word-by-word captions. Pair text with zooms and sound effects for a produced finish. See the TikTok creators, YouTubers, podcasters (lower thirds for guests) and SaaS demos workflows, the text and overlays docs, or the AI video editor tour.

Over Valmera's MCP server, these are add_text, add_kinetic_text, set_typography_scene, set_text_motion, add_title_card and add_vector_graphic.