Precise text targets
Mark single words, full sentences, or wrapped lines. Reassign a selection without losing your animation settings.
Select text, give it an animation, and play the result.
No animation cues yet. Select text and add one to begin.
Resolution applies to PNG and WebM. Video export runs in approximately real time and requires browser MediaRecorder and canvas recording support. WebM availability depends on the browser. Transparent overlay PNG contains only the animated strokes at the current playhead—not the text.
Animate marker strokes, underlines, boxes, circles and animated brackets over the exact words that matter. Create polished text callouts for videos, teaching and presentations—with no subscriptions or uploads.
Mark single words, full sentences, or wrapped lines. Reassign a selection without losing your animation settings.
Animate highlights sequentially or together. Scrub, preview, copy cues, and manage each drawing speed.
Save frames, transparent annotation overlays, WebM recordings, projects and CSV cue sheets—all locally.
Yes. The studio segments each animation across the selected wrapped lines. Long text may need smaller font sizing.
No. Animations are generated deterministically by the browser's canvas engine. No remote processing is needed.
Save the project as JSON and reopen it in the studio. The project includes text, scene settings and all animation cues.
Video recording requires canvas capture and a browser-supported MediaRecorder codec. Use PNG or save the project when your browser lacks WebM encoding.
Preparing your animation…
0%Paste text, select a word or phrase, and add a marker, underline, box, or circle cue. Refine each cue’s order, color, and timing, then preview how it follows wrapped lines. Export a PNG, transparent stroke-only PNG, cue CSV or JSON, or a silent WebM when browser recording is supported.
Authored example: In a product notice, underline “early access” and circle the release date. Check both cues after the text wraps before exporting.
Limits and privacy: Phrase suggestions use deterministic rules, not AI. The editor accepts up to 2,800 text characters, 80 cues, and a 60-second timeline. System fonts and Canvas text wrapping vary; WebM is silent and browser-dependent, while transparent output is a still image rather than alpha video.