AI avatar explainer videos for SaaS: how they work and what they cost

An AI presenter can explain your product on camera without filming anyone. What an avatar video is, when it beats motion graphics, how to write the script, and the cost.

The short answer: an AI avatar explainer is a video where a realistic AI presenter speaks your script on camera, so you get a person explaining your product without filming anyone. It works best for founder messages, onboarding and sales outreach. In Relo you pick one of ten presenters, write the script, and motion graphics are built around them.

  • Best for: "here's what we built and why", onboarding, personal sales videos, feature explainers.
  • Less suited to: fast, music-led social ads, where motion graphics usually hold attention better.
  • Cost in Relo: 35 credits for the video + 35 for the avatar + 10 per render, so 80 credits for one finished size.

What is an AI avatar explainer video?

It is a video in which a digital presenter, generated by AI, talks to the camera and says your script. Their lips, face and voice are synthesised; you never set up a camera, lights or a microphone.

In Relo the presenter is part of a designed video, not a talking head on a plain background. Motion graphics (titles, product screenshots, callouts) appear around them as they speak, the way a produced explainer would cut between a host and the product.

When should a SaaS company use an avatar instead of motion graphics?

Use an AI avatar when…Use motion graphics when…
A human face builds trust, as in a founder note or a sales messageThe product can be shown rather than explained
The idea needs explaining in sentences, such as a new workflow or a pricing changeYou want a fast, punchy ad for a social feed
You want onboarding to feel personalThe video should work with the sound off
You'd otherwise have to film yourselfYou don't want a person in the video at all

Many teams use both: an avatar for the onboarding and sales videos, motion graphics for launch and ad cuts.

How do I make one?

  1. Choose a presenter. Relo has ten, each with their own voice. Choose the one who suits your audience; a developer tool and a wellness app usually want different presenters.
  2. Write the script. This is what the presenter will say, word for word (see the tips below).
  3. Attach up to two images if you like, such as a product screenshot or your logo, to appear in the motion graphics.
  4. Review the draft and ask for changes to the graphics, captions or timing, in plain words.
  5. Render in 16:9 for your site and email, or 9:16 for social.

The presenter does the speaking, so there's no separate voiceover to add.

How do I write a script an avatar can deliver well?

  • Write for the ear. Short sentences. One idea each. Read it aloud; if you run out of breath, split the sentence.
  • Open with the viewer's problem, not your company name.
  • Explain one thing. "How to connect your bank in two minutes" works better than a full product tour.
  • Say numbers the way you'd speak them: "two minutes", not "2 min".
  • End with one next step: "Click Connect bank at the top of your dashboard."
  • Aim for 60–150 words. That's roughly 30–60 seconds of speech.

An example opening:

Hi, I'm Kira from Ledgerly. If you've ever chased a client for an invoice three times, this is for you. In the next minute, I'll show you how Ledgerly sends the reminders for you.

What does an AI avatar video cost?

Filming a presenter means booking a person, a camera and an edit, which often costs more than the rest of the video. With Relo it's priced in credits:

ItemCredits
The video35
AI avatar+35
Each change you ask for5
Final render, per size10
One finished video, one size80

Captions, music and your brand kit are free. The avatar is in beta. See pricing for credit packs.

Are AI avatar videos honest to use?

They are, if you don't pretend the presenter is a real employee. Some teams say so in the video or description ("presented by an AI host"). Never use an avatar to impersonate a real person, and keep the claims in the script as accurate as you would in writing.

FAQ

Can the avatar speak my script word for word?

Yes. The presenter speaks the script you write; you can edit it and regenerate.

Can I use my own face or voice?

Relo's avatars are its ten built-in presenters. If you want your own voice without a presenter, use a motion-graphics video with your cloned voice as the narration.

How long can the script be?

Up to 4,000 characters, which is far more than a good explainer needs. Most work best at 30–60 seconds.

Is an avatar better than a cartoon presenter?

They suit different moods. Relo also has a presenter mode with cartoon hosts for concept explainers and product reviews, which feels lighter and more playful than a realistic avatar.

Vraj Patel, Founder, Relo. Founder of Relo. He builds the product and writes these guides from what he sees making product videos with AI every day.