Talking Head Tutorial — Demo
The 20-second showcase: one screen recording, the Guide cycling square → bubble → cutout in the corner, content audio on the timeline.
The course and walkthrough layout: your screen recording plays full-bleed while the Guide talks it through from a corner — rounded square, circle bubble, or a background-removed cutout — with a title card and closing beat built in.
what you get
Put yourself on screen over anything — a screen recording, slides, gameplay, a product demo, or footage you shot. One clip of you talking becomes a floating Guide in three interchangeable looks, and switching between them is a single ask to your agent.

A rounded-square facecam in the corner — the look viewers know from tutorials, courses and walkthroughs. Your take is auto-cropped square, headroom preserved.

A round bubble floating over your content — the screen-share, reaction and commentary look. Same footage, different frame; swap anytime.

You with the background removed, standing right on the content — the modern presenter look for launches, promos and talks. Real video matting, produced locally — not a CSS trick.
how it works
Any screen capture works. It plays full-bleed as the content, and its sound lands on the timeline where the mix belongs.
Give your agent the webcam take. It becomes the Guide in the corner — square, bubble, or cutout, bottom-right by default, any corner by asking.
The intro card and closing card re-anchor to your length automatically. Render an MP4 on your own machine — no credits, no watermark.
use it for
The format students expect: the lesson on screen, the teacher present. Batch a series by swapping the recording and narration per lesson.
Show the real product while a person carries the story. The presence of a face measurably holds attention through setup steps.
Answer the questions your team repeats — one screen recording with a face, sent as a link. Update the video when the UI changes by re-recording only the screen.
Finished projects that ship inside the template — open one in the app to see how it was built.
A talking head tutorial is a screen recording with the presenter’s face visible while they explain it — the format most courses, walkthroughs and screencasts use. The face usually sits in a corner over the screen content. This template ships that layout finished: your recording plays full-bleed, the Guide talks it through from the corner, and an intro and closing card frame the video.
Drop both files into the project and ask your agent. The screen recording becomes the full-frame content, its audio lands on the timeline, and your webcam take becomes the Guide in the corner — square-cropped automatically for the boxed styles. There’s no manual masking or keyframing; describing the result is the workflow.
Yes — that’s the cutout style. It uses real video matting produced locally by the studio’s effects tooling, not a green screen or a CSS trick, so an ordinary webcam take works. When matting isn’t available on your machine, the square and bubble styles are the designed fallback — same footage, boxed instead of cut out.
Anything where a person needs to carry the explanation: course lessons, product walkthroughs, onboarding videos, support answers, code reviews. Viewers hold attention longer when a face is present, and the corner placement keeps the actual content readable. The three styles let the same footage read as a course (square), a call (bubble), or a stage presence (cutout).
Yes — bottom-right is the default because it covers the least UI in most recordings, and any corner is one ask away. The cutout style supports the bottom corners, since a silhouette hanging from the top edge reads wrong; the boxed styles go in any of the four.
Yes — ask your agent for captions and it transcribes your voiceover, times each line to the audio, and places them on the video. Position, size and style are also one ask away, and the same goes for other edits: retimes, corner changes, a music bed. Describing what you want is the workflow.
Yes. The starter opens as a 20-second showcase so you can judge the three styles immediately, and retiming to your recording’s real length is one ask — the intro card, the Guide, and the closing card re-anchor themselves. There’s no upper limit from the template; renders run on your own machine.
more templates
CoAnimator is a full agent-driven video studio — timelines, voiceover, effects and rendering, all on your machine. Your agent does the editing: ask for captions and they arrive transcribed from your voiceover and timed to it, ask for a retime, a style switch, a music bed — it just happens. Every template is a working project to start from: