MakeAIVideo
Start trial

Make AI talking avatar videos

Pick a presenter from the library or upload a reference photo, paste your script, and MakeAIVideo lip-syncs an AI character to a natural voice with word-by-word captions in 1080p. Selling? See AI spokesperson video. Running ads? See AI UGC video. Already filmed it? Character swap replaces the person in your own footage.

Start 7 day free trialSee pricing

7-day free trial, 350 credits·Plans from $9/mo · Cancel anytime

Join 690+ creators making AI video
Talking Avatar example: A lip-synced AI presenter from your script, made with MakeAIVideo

An AI presenter lip-synced to a script.

Watch this example full size

No filming, no actor, no studio

An AI talking avatar on camera. No camera.

An AI talking avatar is a presenter who reads your script to camera without anyone being filmed. Pick an AI character (or upload a reference image to generate a custom one: one photo is enough to make an AI avatar from), paste your script, and MakeAIVideo lip-syncs the presenter to a natural AI voiceover with word-by-word captions, then renders an MP4. A presenter video without filming, hiring, or a studio, and because the character is AI-generated rather than cloned from a real person, there's no actor likeness to license. Prefer scenes over a face? Script to video narrates the same words over AI footage.

Lip-synced delivery.
The character's mouth is timed to the voiceover.
An avatar from a photo.
Use a library face, or add one to five reference photos and the wizard keeps that face in every portrait.
Natural AI voice.
Match the voice to the character from a curated library of narrators.

How it works

How an AI talking avatar video works

  1. Step 1

    Pick an AI character

    Choose a presenter from the curated library, or make an AI avatar from a photo: in the character wizard one reference image is enough on its own, and up to five lock the face. Pick a portrait, give it a voice and save it. No real person is filmed, cloned, or recorded; the character is AI-generated and synthetic from the start. Would rather appear yourself? Choose Upload your video instead and add a 30 to 60 second landscape clip of you talking.

  2. Step 2

    Paste your script and pick a voice

    Drop the words you want said into the script box, or use Write with AI to draft them from a topic. The counter estimates the runtime as you type: about 150 words makes a 60 second talking head. Choose a voice from a curated library of natural narrators that matches the character's apparent age, tone, and pace before generating.

  3. Step 3

    Generate and download

    MakeAIVideo lip-syncs the presenter to the voiceover, adds word-by-word captions, and renders a 1080p MP4 sized for the aspect ratio you picked. Open in the editor to tweak, or post the default output.

Under the hood

How a photo becomes a talking avatar with lip sync

The Talking Avatar screen has two ways to put someone on screen. Both start from your script and a voice; they differ in where the face comes from and what the lip sync works on.

Pick a character

A still becomes a talking presenter.

  1. 1Start from a library presenter or a character you made from one to five photos in the character wizard.
  2. 2A library presenter starts from its own portrait. For a character you made, MakeAIVideo makes a still for each part of the script, framed from mid-chest up, face to camera, mouth closed and hands out of frame.
  3. 3Our AI lip-sync model takes that still and the voice track and animates the face, mouth and head to the words.

Best for: Course lessons, explainers and a recurring presenter you never have to film.

Upload your video

Your own footage, new words.

  1. 1Upload a 30 to 60 second landscape clip of you talking neutrally, well lit, against a simple background.
  2. 2The new script is voiced, then the mouth in your clip is re-synced to it, so the person on screen is really you.
  3. 3Switch on Add B-Roll and you speak the hook while real stock footage plays during the body.

Best for: Founder updates and ads where viewers should see the real person.

Want a photo to move without a script, or a short lip-synced clip from a single still? That is image to video, a separate mode that makes one 5 or 10 second clip.

The Talking Avatar screen

Make a talking avatar AI video in your browser

Pick the character, paste the script, generate. The talking avatar AI screen has two starting points: choose a presenter you have saved, or Upload your video to lip-sync a clip of yourself instead, then write or paste up to 5,000 characters of script, pick the voice and the aspect ratio, and the finished clip lands in your library.

Pick a character, or upload your video.
Any presenter in your library, or a 30 to 60 second clip of you talking.
Your script.
Read word for word, with captions timed to the voice.
Generate video.
A 1080p MP4 in 9:16, 16:9 or 1:1.

Lip sync

What makes talking avatar lip sync look natural

Most of what makes a talking avatar look real is decided before you press generate: the face you give it, the words you give it and the voice you pair them with.

Frame the face large and front-on

The lip-sync model animates whatever still it is given. One person, chest up, looking into the lens, with nothing in front of the mouth gives it the most to work with. A small figure in a wide shot gives it the least.

Start from a closed, relaxed mouth

A reference photo with an open mouth, a big grin or a hand near the face fights the lip sync. A neutral expression with the mouth at rest reads best once it starts talking.

Light the face evenly

If you're uploading your own reference image, choose a frontal shot with even lighting on the face. Side-lit or backlit photos lip-sync less accurately than well-lit ones.

Film the upload the way the app asks

For Upload your video, record 30 to 60 seconds in landscape, talking neutrally, well lit, against a simple background. Big head turns and hands across the face give the re-sync less to hold on to.

Write for the ear, not the page

Natural speech lip-syncs better than tongue-twisters or dense technical jargon. Read the script aloud once before generating; if it trips you up, it will trip the model up too.

Pace the script to the counter

The script box estimates the runtime as you type, at about 150 words a minute. Short sentences give the voice natural pauses, and the presenter's face settles in them instead of talking non-stop.

Match the voice to the face

A young face with an elderly voice (or vice versa) breaks the illusion. Pick a voice that fits the character's apparent age and tone before generating.

Check it muted, then with sound

Watch the first render with sound on to catch a line that sounds robotic. Watch it again muted, the way much of a feed is watched, and make sure the captions carry the point alone.

AI avatar generator

Four AI avatar video jobs, and which a talking avatar does

The same search covers products that work very differently. Here is which ones Talking Avatar does, and the one it doesn't.

Read a script to camera

This is Talking Avatar

Pick a character, paste the script, choose a voice. The presenter is lip-synced to the voiceover and captioned, and the finished MP4 lands in your library.

Look like you

From a photo or your own footage

Make a character from a single photo of yourself in the character wizard, or choose Upload your video and add a clip of you talking. The voice comes from the library: MakeAIVideo doesn't clone voices.

Be one scene in a longer video

Add B-Roll

With your own uploaded video, switch on Add B-Roll: the avatar speaks the hook and real stock footage plays during the body, the classic ad format. No presenter at all? Script to video narrates over scenes.

Hold a live conversation

Not what we make

A real-time avatar that listens and answers face to face inside a product is a different kind of tool, billed by the conversation minute. MakeAIVideo renders finished video files. It doesn't make profile-picture avatars either.

Decide

Talking avatar or film yourself?

A talking avatar takes the camera out of the job. That is a big saving for some videos and the wrong call for others.

Setup
Talking avatar: A browser tab. No camera, lights, mic or quiet room.
Filming yourself: A camera or phone, light on your face, a clean mic and a quiet room.
A fluffed line
Talking avatar: Edit the script and generate again.
Filming yourself: Another take, and another trip through the edit.
Changing one line later
Talking avatar: Change the words in the script; the presenter, voice and framing stay the same.
Filming yourself: Re-film, and match the lighting, outfit and framing of the original.
Another language
Talking avatar: Write the script in that language and the voice reads it.
Filming yourself: Learn the lines, hire a native speaker, or dub.
Same look every week
Talking avatar: One saved character comes back identical in every video.
Filming yourself: Depends on the day, the hair and the light.
Where filming wins
Talking avatar: An AI presenter can't hold or demo a physical product, and it can't give a real customer's opinion.
Filming yourself: Your real personality, hands-on demos, live reactions and genuine testimonials.

Already have footage and want someone else in it? AI character swap replaces the person in your own clip and keeps the movement.

After it renders

Every talking avatar clip in one place

Renders land in your library, ready to download, re-edit, or post with captions baked in. Restyle the captions or change the background music in the editor without generating again. See pricing or start from one line with prompt to video.

Every clip in My Videos

See what is still rendering and what is ready, then reopen, download or post any presenter clip.

Edit after generating

Restyle the captions, or click any word in the transcript to fix a typo without re-recording the voice.

Export anywhere

Download a 1080p MP4 with SRT or VTT caption files, or copy a link anyone can download it from.

Post it from here

Send the clip to TikTok, Instagram, YouTube and eight more platforms, now or on a schedule.

Connect your AI agent

Ask your AI assistant for a talking avatar video

Connect Claude, ChatGPT, Grok or any MCP client to your MakeAIVideo account, paste the script into the chat and name the presenter. It makes the clip in your library and posts it if you ask, with the same credits and from the same balance as the app.

Connect your AI assistant.
Add MakeAIVideo to Claude, ChatGPT, Grok, Claude Code, Cursor or any app that supports MCP with one address. You sign in once; there is no API key to paste.
Ask in plain English.
“Make a 40 second vertical talking avatar video with Maya reading this product update, with captions.” It checks the cost before it spends anything.
It makes the video.
A finished MP4 with the script, voiceover, scenes, captions and any music, in your library to edit or download.
And posts it if you ask.
To the accounts you have connected, now or on a schedule, across 11 platforms.
Start free trialConnect any AI agent

MCP, API and CLI use the same credits and the same balance as the app.

Connector address

mcp.makeaivideo.ai

Works with

Signed in? Connect an AI assistant

All included

Everything a talking avatar video needs

No filming, no actors, no studio. The whole presenter pipeline runs from your script.

Character library.
Pick a ready-made AI presenter or generate one from a reference image.
Lip-sync engine.
Mouth movements timed to the voiceover.
Studio AI voices.
A curated library of natural narrators to match the character.
Word-by-word captions.
A choice of caption styles, timed to the voice and burned into the export.
Vertical, square or widescreen.
9:16, 1:1 or 16:9, one ratio per export, captions adjust to each.
Watermark-free.
Clean exports, yours to post anywhere.

Specs and limits

AI talking avatar specs, limits and costs

What goes in, what comes out, and what Talking Avatar doesn't do, in one place.

Presenter
A library character, a character made from 1 to 5 photos, or Upload your video (a 30 to 60 second landscape clip)
Lip sync
Our AI lip-sync model animates a picked character from a still and the voice track
Script
Up to 5,000 characters, typed, pasted or drafted with Write with AI; about 150 words is a 60 second video
Voice
12 curated studio-quality AI narrators, filterable by gender, accent and tone; no voice cloning or audio upload
Language
The voice reads the language the script is written in; 24 output languages are supported
Captions
Word-by-word, timed to the voice; restyle them or fix a word in the editor afterwards
B-roll and music
Add B-Roll with real stock footage on the upload path; background music you can change in the editor
Output
1080p MP4, no watermark, with SRT and VTT caption files; 9:16, 1:1 or 16:9, one ratio per render, re-export another from the editor
Cost
Credits per second of finished video, all in; a talking presenter is about 19.5 credits a second (30 seconds is about 584), lip sync on your own uploaded clip about 8 a second (30 seconds is about 239), and the exact cost shows before you generate
Getting started
A 7-day free trial with 350 credits on any plan, card required; plans from $9/mo, and credits you buy never expire

No voice cloning

Every presenter is voiced by a curated narrator. There is no audio upload.

No end card

The video ends on the presenter's last line; add a call to action in the script.

Not live

Talking Avatar renders a video file. It doesn't hold real-time conversations.

Use cases

Talking avatar videos you can make with one AI presenter

One AI video presenter, saved once, can front every kind of script below with the same face and voice.

A presenter in a grey sweater speaking to camera in a bright study, a frame from a course lesson

Course lessons

Open every lesson with a consistent instructor: clean audio, focused framing, no background noise. When a module changes, edit that lesson's script and generate it again.

A presenter in a navy t-shirt speaking to camera in a bright office, a frame from a product walkthrough

Product walkthroughs

Narrate a feature tour or onboarding flow with a friendly face. No coordination with engineering or design; re-render any time the product changes.

A presenter with short grey hair speaking to camera against a pale blue backdrop, a frame from an explainer

Explainers

One idea, one presenter, 60 to 90 seconds. A face to camera holds attention on a concept that has nothing to film, from a pricing change to how a process works.

A presenter in a light blue blazer speaking to camera in a bright room, a frame from a translated presenter video

Multilingual versions

Translate the script, keep the presenter. Each language is its own render with the same face, framing and captions, so a Spanish or German version doesn't mean a new shoot.

A presenter in a grey knit sweater speaking warmly to camera in a living room, a frame from a personal brand video

Personal brand

Build a recognisable on-camera presence without filming. The same AI presenter, every week, with a consistent look and voice across every clip.

A presenter in a white blouse speaking to camera in a clean office, a frame from a help centre answer video

Support answers on video

Turn the questions your support team answers every day into short, calm presenter clips for the help centre. When the product changes, edit the script and re-render the answer.

If you want generated scenes rather than a presenter on camera, script to video reads your words over footage instead.

Best practices

Tips for talking avatar scripts people watch to the end

Aim for 30 to 90 seconds

Talking-head videos work best in the 30 to 90 second range. A short clip is easier to watch to the end, especially on social, so save the long read for a second video.

Front-load the hook

The first three seconds decide whether the clip survives the scroll. Open with the most interesting line, not a generic greeting.

Keep the first videos simple

Start with a 60 second explainer or a short how-to: a 5 second hook, 45 seconds of body, a 10 second call to action. Save the ambitious formats until you know how your presenter reads.

Stay with one presenter

An avatar starts to feel familiar only after viewers have seen it a few times, so consistency matters more than finding the perfect face. Save one character that fits your brand and use it in every video.

Don't oversell with an AI presenter

AI avatars work best for educational, explainer, and informational content. Overly emotional or hard-sell delivery from an AI character reads as inauthentic faster than from a human.

Never cast the avatar as a customer

An AI presenter speaking as your brand is fine. An AI presenter playing a satisfied customer who doesn't exist is a made-up endorsement: the FTC's Endorsement Guides Q&A says an endorsement must reflect the endorser's honest opinion, and you can't describe experience with a product nobody tried.

More ways to make AI avatar and presenter videos

AI Spokesperson VideoA business presenter for company videos.AI UGC VideoCreator-style ads for paid social.AI Influencer GeneratorOne persona posting across channels.AI Character SwapSwap the person in your own footage.Image to VideoAnimate a photo into a 5 or 10 s clip.

Talking avatar guides and comparisons

HeyGen vs Synthesia 2026Head-to-head on avatar quality, voices and price.Compare HeyGen vs SynthesiaBest Synthesia alternatives 20268 Synthesia alternatives tested, with real 2026 pricing.See the Synthesia alternativesBest HeyGen alternatives 202610 talking-head and personalised-outreach tools tested.See the HeyGen alternativesTikTok video generatorVertical videos sized for the For You feed, with captions and posting built in.Make TikTok videos with AIFaceless YouTube channelRather keep any face off camera? Narrated videos over scenes, start to finish.Start a faceless YouTube channelPrompt to videoOne line of text becomes a narrated video with scenes instead of a presenter.Try text to video from a prompt

Pricing

Talking avatar pricing: plans priced in credits

Talking Avatar is on every plan, with 1080p downloads, no watermark and posting to 11 platforms. A talking presenter is about 19.5 credits a second, so 30 seconds is about 584. Start with a 7-day free trial, then plans from $9/mo.

Lite

The easy start. Every tool, one creator.

$9/mo

or $99 a year, 1 month free

600credits every month

Use them on any tool below

Start 7 day free trial

7 days free with 350 credits. Cancel before day 7 and you won't be charged.

  • Post and schedule to 2 social accounts
  • Videos up to 10 minutes
  • 1080p export, never a watermark
  • 1 user

Every tool included

Solo

For one creator posting every week.

$19/mo

or $190 a year, 2 months free

1,100credits every month

Use them on any tool below

Start 7 day free trial

7 days free with 350 credits. Cancel before day 7 and you won't be charged.

  • Post and schedule to 4 social accounts
  • Videos up to 10 minutes
  • 1080p export, never a watermark
  • 1 user

Every tool included

Most popular

Plus

Long-form, your branding and a team of three.

$29/mo

or $290 a year, 2 months free

1,750credits every month

Use them on any tool below

Start 7 day free trial

7 days free with 350 credits. Cancel before day 7 and you won't be charged.

  • Post and schedule to 6 social accounts
  • Stock-footage videos up to 40 minutes
  • Brand Kit: your logo on every video
  • 3 team seats with admin and member roles

Every tool included

Pro

For teams and channels running at volume.

$49/mo

or $490 a year, 2 months free

3,000credits every month

Use them on any tool below

Start 7 day free trial

7 days free with 350 credits. Cancel before day 7 and you won't be charged.

  • Post and schedule to 10 social accounts
  • Stock-footage videos up to 40 minutes
  • Brand Kit: your logo on every video
  • 10 team seats with admin and member roles

Every tool included

Run out in a busy month? Add credits any time, from $5. Credits you buy never expire. See pricing

Every plan includes

  • Captions, music and studio-quality voices, free on every video
  • Full HD 1080p export, never a watermark
  • Post and schedule to 11 platforms
  • Video in 24 languages
  • API, MCP server and CLI for ChatGPT, Claude or any AI agent

AI talking avatar FAQs

What people ask before they make their first presenter video.

Something else? Visit the help center

What is a talking avatar video?

A talking avatar video features an AI character (the presenter) lip-synced to an AI voiceover of your script. MakeAIVideo generates it without a camera, filming, or a real actor.

How do I make an AI talking avatar video?

Open Talking Avatar and choose Pick a character for a presenter from your library, or Upload your video to use a clip of yourself. Paste or write up to 5,000 characters of script, pick a voice from the curated library and an aspect ratio, then generate. The lip-synced MP4 lands in your library with word-by-word captions.

What is an AI video presenter?

An AI video presenter is a synthetic person who reads your script to camera, so you get a presenter-led video without filming anyone. On MakeAIVideo that is Talking Avatar: the presenter is lip-synced to a natural AI voice and captioned. The same presenter can front a product walkthrough, a course intro or an explainer.

How does AI avatar lip sync work?

For a picked character, MakeAIVideo voices your script, makes a still of the character framed from mid-chest up, and our AI lip-sync model animates that still to the voice track, moving the mouth, face and head with the words. With Upload your video, the mouth in your own clip is re-synced to the new voiceover instead.

How realistic are AI presenters?

Realistic enough for lessons, explainers and walkthroughs when the input is good: a large, front-on, evenly lit face, a conversational script and a voice that fits the face. A side-lit photo, a small figure in a wide shot or a script full of jargon all show up in the result. Nobody can promise a lip sync that is flawless in every frame, so watch each render before you post it.

Is there a talking avatar AI app I need to download?

No. MakeAIVideo runs in your browser, so there is nothing to install. You can also ask Claude, ChatGPT, Grok or any MCP client connected to your account to make the talking avatar video for you.

Can I use my own character?

Yes. Upload a reference image and MakeAIVideo generates a consistent AI character from it. Voice selection comes from a curated library of natural AI voices.

Is this an AI person generator?

In the sense that matters, yes. Upload a photo and MakeAIVideo creates an AI character based on it, then lip-syncs that character to your script. The person is an AI likeness, not your raw photo: nobody is filmed or recorded, and the same face comes back in every video.

Is it a real person?

Not when you pick a character. The presenter is an AI-generated character, not a real person, and the voice is AI-synthesised. That keeps you clear of likeness and rights issues. The one exception is your choice: with Upload your video, the person on screen is whoever is in the clip you upload.

Can I make an AI avatar from a photo?

Yes. Open the character wizard, add a reference photo and it generates portraits that keep that face; one photo is enough, and you can add up to five. Pick the portrait you like, choose a voice, name the character and save it. From then on it is in your Talking Avatar picker for every script.

Can I make an AI avatar that looks like me?

Yes, two ways. Make a character from a photo of yourself, or choose Upload your video in Talking Avatar and add a 30 to 60 second landscape clip of you talking neutrally, well lit, against a simple background; each new script is lip-synced onto that footage. Either way the voice comes from the curated library, because MakeAIVideo doesn't clone voices.

Can I use my own voice?

No. MakeAIVideo doesn't clone voices and doesn't take an audio upload, so every talking avatar is voiced by one of the curated studio-quality AI narrators. Pick the one closest to the tone you want and use it in every video so your presenter always sounds the same.

Can the avatar speak other languages?

Yes. Write or paste the script in the language you want and the voice reads it in that language, with captions to match. MakeAIVideo supports 24 output languages. Each language is its own render, with the same presenter if you pick the same character.

Can I add B-roll to a talking avatar video?

Yes, with Upload your video: switch on Add B-Roll and you speak the hook while real stock footage illustrates the body. B-roll is stock footage, not clips you upload. After the render, the captions and the background music can be changed in the editor.

What is the difference between an AI talking head and an AI avatar?

Mostly the angle you look at it from. A talking head describes the framing: a chest-up presenter speaking to camera. An AI avatar describes the synthetic person doing the talking. A talking avatar video is both, and it is made the same way: a character, a script, a voice and lip-sync.

What is the difference between an AI avatar and a deepfake?

Consent and honesty. An AI avatar is a character generated for the job, or footage of someone who agreed to it, speaking as your brand. A deepfake puts a real person's likeness to words they never agreed to say. Only upload photos or video of yourself or of people who have given you permission, and never present the avatar as a real customer.

Do I have to label an AI avatar video on YouTube?

It depends on what the video shows. YouTube's altered or synthetic content policy asks creators to disclose realistic AI content that makes a real person appear to say or do something they didn't. A lip-synced clip of you reading a new script, or an avatar made from a real person's photo, can fall under that, so tick the disclosure when it does. Using AI only for the script or captions doesn't need it.

Can I have a face-to-face conversation with an AI avatar?

Not on MakeAIVideo. Talking Avatar renders a finished video file from a script. A real-time avatar that listens and answers face to face inside a product is a different kind of tool, usually billed by the conversation minute.

Is there a free AI avatar generator?

Every plan starts with a 7-day free trial with 350 credits, full quality and no watermark; a card is required to start it. A 15 second talking presenter video is about 310 credits. After the trial, plans start at $9 a month for 600 credits. See pricing.

Does it include captions?

Yes. Word-by-word captions are generated and timed to the voiceover, in your choice of caption style.

How much does it cost?

Each video costs credits, all in: voice, lip sync, captions and export. A talking presenter is about 19.5 credits a second, so a 30 second video is about 584. Lip sync on your own uploaded clip is about 8 credits a second, so 30 seconds is about 239. A talking photo is about 22 credits a second of speech, so 10 seconds is about 216. The screen shows the exact cost before you generate. Plans start at $9 a month for 600 credits; making presenter videos every week? Plus is $29 a month for 1,750 credits.

How long can the script be?

Up to 5,000 characters. Talking-avatar clips work best in the 30 to 90 second range, which is easier to watch to the end on social. Longer scripts are supported, but render time and credit cost scale with the spoken duration.

How long does it take to render?

A few minutes for a typical clip, and longer for a longer script. The render lands in your library when it's ready.

What aspect ratios are supported?

9:16 vertical (TikTok, Instagram Reels, YouTube Shorts), 16:9 landscape (YouTube), and 1:1 square (Instagram feed). Pick the ratio before generating; each render is one ratio, and you can switch ratio in the editor and export again.

Can I use the clips commercially?

Yes. Videos you generate are yours to use commercially, for ads, product pages, social posts, and client work. MakeAIVideo terms of service.

What happens when I run out of credits?

Nothing is lost: your account and finished videos stay exactly as they are. Your plan credits refresh on your next billing date. Run out sooner? Add credits any time, from $5; credits you buy never expire.

Put a talking avatar on your next script

Pick a presenter, paste your script, and get a lip-synced 1080p MP4 with word-by-word captions.

Start 7 day free trial