How to make faceless YouTube videos with AI
By Neil, founder of Lantaw. Last updated: 2026-10-25
You can make a faceless YouTube video with AI in four moves: write a script, let an AI voice read it, add footage that follows the story (AI b-roll, or an AI presenter), and burn in captions. In Lantaw those steps happen in one place, and the finished 1080p video can be published straight to your channel.
This guide walks through each step, what makes a faceless video worth watching past its first few seconds, and what it costs.
What is a faceless YouTube video?
A faceless video does not show the creator on camera. The viewer hears a narrator and sees footage that matches what is being said. The most common formats are narrated storytelling (history, true crime, mystery, explainers, news recaps) and videos led by an AI presenter, a consistent on-screen character made from a single photo.
How to make a faceless YouTube video from a script with AI
Pick a topic worth a whole video
Choose a narrow niche and a specific angle, not a broad subject. In Lantaw, Topic Finder searches the web for your niche and audience country and returns eight video ideas, each with a hook, an angle and its sources.
Write the script for the ear
Write the way people speak: short sentences, one idea per line, and a first sentence that makes a promise or raises a question. You can turn a Topic Finder idea into a spoken script in one click, or paste your own. The video runs exactly as long as the script takes to say, roughly 45 to 60 seconds for every 150 words, depending on the voice.
Choose the look: b-roll or an AI presenter
For a faceless channel, choose a narrated video with b-roll and no presenter. An AI director writes one shot for each scene, so the footage follows what the narration says, set in your audience’s country. If you want a recurring face, save an AI avatar from one photo plus a voice and reuse it on every video.
Pick a voice and a format
Pick a ready-made natural voice. Choose 16:9 landscape for standard YouTube videos or 9:16 portrait for Shorts and other short-form feeds.
Add captions
Captions are timed to the spoken words and burned into the video in five styles, in a colour you choose.
Generate, review and publish
Longer scripts are cut into clips and take longer to make, so start with a short one. When it is done, download the 1080p video, or connect up to 10 YouTube accounts and publish with a title, description and thumbnail, now or on a schedule. Lantaw can label the upload as AI-generated in one click.
What makes a faceless video worth watching
These are craft habits, not guarantees. No tool can promise views.
- Open with the most interesting thing, not a greeting. The first sentence of the script is the hook.
- Make one promise per video and make every scene serve it.
- Be specific. Names, numbers and concrete details hold attention better than general statements.
- Match the footage to the line. B-roll that shows what is being said feels deliberate; unrelated footage feels like filler.
- Vary the rhythm: mix short and long sentences, and leave a beat before the key line.
- Write a new script for every video instead of reusing one template.
How much does a faceless AI video cost?
Lantaw plans start at $20 a month (Spark). Every plan comes with monthly credits, and credits are spent by the second of finished video: about 13 credits per minute for generated b-roll and 33 for an AI presenter. You are charged for the finished video’s real length, and a video that fails gives its credits back. New accounts get 50 free credits to use within 7 days, about 3 minutes of video at the default settings. Videos made on the trial carry a small Lantaw watermark. No card needed.
Will YouTube monetize AI faceless videos?
That is YouTube’s decision, and no tool can promise it. YouTube’s monetization rules do not allow mass-produced or repetitive videos made from a template with little variation, so write your own script for each video, change the subject and angle, and check YouTube’s current Partner Program policies. Label AI-generated videos as such when you publish.
Faceless YouTube video FAQ
Do I need to show my face to run a YouTube channel?
No. A faceless channel uses a narrator and footage that matches the script. You can also use an AI presenter, a character made from a single photo, so you never appear on camera.
Can I use my own voice?
No. We removed the voice cloning feature, so every video is read by one of the ready-made natural voices. You can still use your own face as an avatar, as long as the face is yours or you have permission to use it.
How long can a faceless video be?
There is no fixed limit. The length is set by how long the narration takes to speak. Longer scripts are cut into clips and take longer to make, and very long videos (30 minutes or more) have had little testing so far.
Can I make YouTube Shorts?
Yes. Choose 9:16 portrait for Shorts and other short-form feeds, or 16:9 landscape for standard YouTube videos. Both render in 1080p.
Is there a free trial?
Yes. New accounts get 50 free credits to use within 7 days, about 3 minutes of video at the default settings. Videos made on the trial carry a small Lantaw watermark. No card needed. After the trial you need a monthly plan, which you can cancel anytime.
Make your first faceless video
Type a script and Lantaw builds the video: narration, b-roll or an AI presenter, and captions.
See plans and pricing About the AI faceless video generator Best faceless YouTube niches for AI video