What is a talking head video?

A talking head video is a video in which one person speaks directly to the camera, usually framed from the chest or shoulders up, with few other visuals. The speaker can be a filmed person or an AI avatar generated from a written script.

Reviewed by Clark Quinn, Chief Learning Strategist Updated October 28, 2026

Example

A bank’s compliance lead needs to introduce a new anti-fraud course. She records a ninety-second talking head video at her desk explaining why the course exists and what has changed since last year. Hearing it from the person accountable for the policy gives the course weight.

For the twelve lessons that follow, the L&D team uses an AI avatar as a talking head presenter, cutting away to slides and worked examples. When the policy changes, they edit the script and regenerate the affected lesson instead of asking the compliance lead to film again.

How it works

A talking head video is defined by its framing and focus: one speaker, looking into the camera, talking to the viewer. There are two ways to make one.

Filmed talking heads

A real person is recorded on camera. A typical setup covers:

• Camera: at eye level, framed from mid-chest up with a little space above the head.
• Microphone: close to the speaker, such as a lapel mic, because viewers notice poor audio faster than imperfect video.
• Lighting: soft and even on the face, from a window, a ring light or a pair of softboxes.
• Prompting: a teleprompter or short bullet notes, so the speaker can look into the lens.
• Background: plain or relevant to the topic, such as an office, a workshop or a branded backdrop.

The footage is then edited, often with captions, name titles and cutaways to slides or demonstrations.

AI-generated talking heads

An AI avatar presents the content instead of a filmed person. The creator writes a script and picks an avatar and a voice. The software turns the script into speech with text-to-speech and animates the face to match. Updating the video means editing the script and generating it again.

Earlier AI avatars moved little more than their lips, which is where the reputation for stiff, static AI talking heads comes from. Newer avatars add hand gestures, head movement and posture changes, so they look closer to a filmed presenter.

Filmed vs AI talking head videos

• Authenticity: a filmed speaker is a real person with their own credibility. An AI speaker is a generated presenter, and viewers should be told it is AI.
• Updates: a filmed video needs a reshoot or an awkward edit. An AI video is regenerated from the edited script.
• Languages: a filmed video needs subtitles, dubbing or a new shoot for each language. An AI video can be generated in each language from a translated script.
• Setup: filming needs a camera, a microphone, lighting and a willing presenter. An AI video needs a script and an avatar.

Choose filmed when the speaker’s identity is the point, such as a leader’s welcome or a personal story. Choose AI for content that changes often or needs many language versions.

Best practices

• Keep it short. Use the talking head where a human presence helps, and switch to visuals for detail.
• Cut away to the subject. Show slides, diagrams or screen recordings when explaining a process or a tool.
• Write for the ear. Short sentences and plain words are easier to follow when spoken than when read.
• Add captions. Subtitles help people who watch without sound or in their second language.
• Look into the lens. Whether filmed or generated, a speaker who looks at the viewer feels more direct than one looking off to the side.
• Keep the setup consistent. Use the same framing, background and presenter across a series, so the videos feel like one course rather than a collection of clips.
• Leave room for overlays. Frame the speaker so titles, captions and slide elements can sit beside or below them without covering the face.

Why it matters for L&D teams

• A human guide. A face and a voice give learners someone to follow through the material, which plain text and slides cannot.
• Simple to produce. A talking head is one of the easiest video formats to plan, film or generate.
• Scales with AI. AI talking heads let teams publish and update presenter-led videos without filming every change.
• Needs support from visuals. On its own, a talking head is weak for procedures and software, so plan the cutaways from the start.