How to make an AI video of yourself talking, in your own face and voice

AI video of yourself: a video in which an AI makes your face and voice say words you never recorded, built from a real recording of you.

AI avatar: a video likeness of a person, generated by AI, that speaks words supplied as a script.

Voice clone: an AI model of your voice that can read new words in your voice.

Persona Video: video messages in your face and your own voice, made on Timeless AI™ from one 30-second recording of you.

Executor Lock™: Idy's patent-pending system of named executors, Trusted Contacts, a verified death report and a sealed snapshot of your Persona that can never be changed.

Yes. You can make an AI video of yourself talking with nothing more than a phone or laptop camera, and the quality in 2026 is good enough that most viewers cannot tell the difference at a glance. Every tool works the same basic way, by learning your face and your voice from a short recording and then generating new video of you saying new words. The tools differ in what they need from you, what they were built for, and who controls your likeness once the video exists. The table compares the main options as of October 2026.

The rival details come from each company's own pages, which we read on October 1 and 2, 2026: HeyGen, Synthesia, Tavus and Creatify. Each one does its own job well. If you need a presenter for a library of training videos, a video-avatar tool is the specialist. If you want video messages in your own face and voice that sit inside an AI version of you, governed by you and passed to the people you name, Persona Video was built for exactly that.

An AI video of yourself works by pairing two models, one that animates your face and one that speaks in your voice, and then matching your lips to the words. The face part learns from footage of you, so the more natural and evenly lit your recording, the more natural every later video looks. The voice part learns from a sample of you speaking and can then read any new script in your voice. The system renders the new words as audio in your voice and moves your mouth to match, frame by frame.

Persona Video works from real footage of you, so your face, hair, lighting and background come from your own recording, and the system animates your mouth to match each new message. That approach keeps every video recognizably you, because nothing about your appearance is invented between one video and the next. Photo-to-video apps take the opposite approach and generate movement from a still image, which is quicker to start and usually less natural to watch. Studio avatars sit at the far end, with professional green-screen shoots that Synthesia says take between one and seven business days to process.

The voice is where most videos succeed or fail, because the people who know you notice a wrong voice at once. A video in your face with a stranger's voice feels wrong to anyone who knows you, while a video in your own voice feels like you even on a small screen. That is why Persona Video takes your face and your voice from the same 30-second recording, and why your Persona speaks with that same voice when the people you choose talk to your Persona.

You can make your first Persona Video in one short sitting, most of which is getting your Persona started, and the recording itself takes 30 seconds. The steps below work in any modern browser on an iPhone, an Android phone or a computer. Your first video is free, with no card.

1. Start your free Persona. Sign up on Timeless AI™ and your free build opens, with 25 memories, unlimited chat with your Persona and no expiry date.

2. Tell your Persona a few memories. Answer a handful of your Persona's questions, by typing or by speaking, so your Persona knows enough about you to compose a message in your own words if you want one.

3. Set up your shot. Sit facing a window or a lamp so your face is evenly lit, hold the camera at eye level, frame yourself from the chest up, and choose a quiet room.

4. Record 15 seconds of your face, quietly. After a three-beat count-in, look at the lens and stay relaxed and silent for 15 seconds, which gives Persona Video the footage of your face.

5. Record 15 seconds of your voice, reading aloud. A short passage appears under the lens, and you read the passage aloud in your normal voice for 15 seconds, which gives your Persona your voice.

6. Confirm your consent. You confirm consent for that exact recording, and Persona Video renders only from the recording you consented to.

7. Write your words or let your Persona compose them. Type what you want to say, up to 500 characters, or let your Persona compose a short message in your own words, and each video runs up to 30 seconds.

8. Watch your video, then make the next one. Your video arrives in your Persona, where you watch the video, delete the video if you change your mind, or make the next message from the same recording.

Recording well takes no equipment beyond your phone, and a few habits make the difference between a good video and a great one. The table lists the habits that matter most, based on what every face-and-voice system needs from its source footage.

You can make any short message that is better coming from your face and your voice than from a text box. The examples below are invented and labeled as such, and each one uses Persona Video exactly as the feature works today. They come from the United States and from the other countries where people search for this guide most.

• A father in Atlanta makes a 30-second birthday message for each of his three children every year, in his own face and voice, written in his own words.

• A grandmother in Toronto makes a message about the recipe her family asks about every holiday, explaining the step everyone gets wrong, in the voice her grandchildren know.

• A university lecturer in Boston makes a short introduction to each week of her course, in her own face and voice, so students meet the person behind the material.

• A career coach in London makes a welcome message that explains how she works, what a first session covers and what to prepare.

• A founder in Bengaluru makes a monthly 30-second update for her team, in her own face and voice, saying what matters this month and why.

• A retired engineer in Munich makes a series of short messages explaining the decisions behind his best-known projects, for the colleagues and family who will ask about them later.

People who think about their own likeness carefully have used AI video and voice for exactly these jobs. Reid Hoffman released a 14-minute conversation with an AI twin built from two decades of his own books, speeches, interviews and podcasts (Popular Science, May 1, 2024), and Kara Swisher sat for about two hours of interview so an interactive twin of her could give 77 answers on CNN on May 16, 2026 (StoryFile). For the family angle in depth, read how to preserve your family's history.

An AI video of yourself costs nothing to try and a few dollars a video after that. On Timeless AI™ your first Persona Video is free, Legacy at $14.99 a month includes 150 seconds of video a month, and Eternal at $29.99 a month includes 390 seconds, with each video up to 30 seconds long. Timeless Pro at $59.99 a month includes everything in Eternal plus your Timeless AI™ Twin. All prices on this page are in US dollars.

Worked example: Marcus, a father of three in Atlanta. Marcus is an invented example, and every price and allowance is real. He wants a 30-second birthday message for each of his three children and a 30-second monthly update for his family, which is 15 videos and 450 seconds a year. His free build covers the first video at $0. On Legacy at $14.99 a month he can make five 30-second videos a month, or 60 a year, for $179.88, which works out to $3.00 a video if he uses the full allowance and leaves him 45 spare videos for whatever comes up. Eternal at $29.99 a month gives him 156 thirty-second videos a year for $359.88, or about $2.31 a video, plus unlimited memories and up to five executors.

The video-avatar tools price for a different job. HeyGen's Creator plan costs $29 a month, Synthesia's Creator plan costs $89 a month, and Synthesia's custom studio avatar add-on costs $1,000 a year, based on the pricing pages we checked on October 1, 2026. Those plans are good value for a marketing team producing training libraries. For a person who wants messages in their own face and voice inside an AI version of themselves, Persona Video costs less and comes with an executor. For every plan detail, read every Timeless AI™ plan and price.

Persona Video is built around you as a person, while video-avatar tools are built around content for an audience. HeyGen, Synthesia and Tavus make presenters for marketing, training, sales and support at volume, and they do that job well. Persona Video makes messages in your face and your own voice inside your Persona, the AI version of you that knows your memories and speaks in your voice, and Persona Video is the only option in this guide that names an executor for your likeness. The table compares the points that matter once a video of you exists.

The last row comes from reading each company's full service terms on October 1, 2026, as part of a review of nine AI twin and avatar platforms and about 85,000 words of contracts. None of the nine said what happens to your clone, your account or your earnings when you die. HeyGen's terms allow free-plan accounts unused for six months to be deleted after notice, and Synthesia's terms say custom avatars cannot be exported and are deleted with the account (Synthesia customer terms, July 20, 2026). For the full platform comparison, read the best digital twin apps compared, and for avatar apps alone, read the best AI avatar apps.

Persona Videos stay inside your Persona and play without a download button, so the file stays with you instead of circulating online where anyone could edit or repost your face. That is a deliberate choice, and it is the same choice a careful person makes about any recording of their own voice and face. Your Persona Video exists to carry your words to the people you chose, under your rules.

The people who did this well made the decision themselves, while they were alive, and the stories that went badly were made by somebody else. The table sorts the best-known cases from five countries by that one question. The pattern is consistent across every culture and every legal system in the list.

ABBA shows the commercial power of a likeness that its owners control. The band recorded once, and the avatars have worked every night since, in a show reported to have cost $175 million to build. The Pelkey case shows the opposite edge. The appeals court said the AI depiction "erases the interpretive distance between the family's belief of what the victim would have said and the victim's own voice and opinions", and a video you make yourself, in your own words, sits entirely on the right side of that line. For who has the authority to make an AI of a person, read AI identity governance, explained.

Yes, when you make the video yourself, on a service that records your consent and lets you control where the video goes. Persona Video records your consent against the exact recording you made, renders only from that recording, keeps your videos inside your Persona without a download button, and lets you delete any video at any time. Your Persona is private to you and the people you choose, and twins are for people 18 and over. The safety question that matters most is what other people can do with your face and voice, and a governed, authorized version of you helps with that question too.

Voice and video fraud aimed at families is real. Fox News reported in June 2026 that the FBI's Internet Crime Complaint Center counted $352 million in AI-related scam losses among Americans aged 60 and over in a year, with voice clones built from social media and voicemail audio used to fake a grandchild in trouble (Fox News, June 11, 2026). Agree a family code word that no recording contains, and tell your family where your authorized videos live, so they always know which face and voice are really you. For keeping control of the recordings and data you share, read what to ask any AI provider about your data and how a personal AI should handle consent and control.

Yes. Making an AI video of your own face and voice, with your own consent, is legal across the United States, the EU, the UK, India, China, South Korea and Japan, and the 2024 to 2026 laws protect you from other people using your likeness. Publishing AI video to the public now carries labeling duties in several places, and making AI video of someone else without consent is a crime in a growing list of countries. The table summarizes the rules that matter for a video of yourself.

Courts have also confirmed that your voice belongs to you. The Beijing Internet Court held in April 2024 that a voice actor's rights cover an AI voice built from her recordings and awarded her 250,000 yuan, about $34,500 (Beijing Internet Court), and the Bombay High Court barred anyone from cloning the singer Arijit Singh's voice, vocal style or likeness without his consent in July 2024 (Columbia Global Freedom of Expression). For the law country by country, read who owns your AI likeness after you die, and for voice and likeness as assets in an estate, read the intangible estate.

Yes, and professionals get the most from pairing Persona Video with a twin that answers in their words. A lecturer, coach, consultant or founder records Persona Video messages in their own face and voice, and Timeless Pro, at $59.99 a month or $599 a year, puts their Timeless AI™ Twin online on their website, through an API and inside AI assistants, answering clients in the first person, in their own words and cloned voice, under their name and photo. The twin carries a badge, tells visitors the truth about being an AI twin, replies in each visitor's language and points booking questions to the professional's booking page. You read every conversation and pause any door with one switch.

Universities and firms have already shown the demand. Harvard Business School charges founders $699 for eight weeks of pitch practice against AI avatars of seven of its professors, and Georgia Tech's AI teaching assistant classes recorded 66% A grades against 62% without the assistant (PYMNTS, August 27, 2026). Those avatars belong to the institutions that built them, while a Timeless AI™ Twin belongs to the professional and passes to the executor the professional named. For the full guide by profession, read professional AI for doctors, lawyers, lecturers and coaches, for eight professions with the numbers worked out, read how to monetize your expertise with AI, and for the twin itself, read Timeless AI™ Twin, the twin that answers your clients as you.

When you die, Executor Lock™ hands control of your Persona to the executor you named, seals your Persona as a perfect snapshot, and applies the rules you set while you were alive. An executor or Trusted Contact you named reports your death with evidence, automated analysis and a human reviewer check the evidence, and a seven-day cooling-off follows in which every executor can dispute the report and you can cancel a false report with proof of life. Our Terms call the lock permanent and irreversible (section 8.3), and nothing in the product can change your personality, your tone or your values after the lock. Your executor can ask Idy to delete your Persona entirely, which gives your family a clean ending whenever they want one.

Robin Williams showed how much a person can decide in advance. His trust barred any use of his name, signature, photograph or likeness for 25 years after his death, with no adverts, no digital insertion into films and no holograms until at least August 11, 2039 (Harvard Journal of Sports and Entertainment Law, March 31, 2015). James Earl Jones made the opposite decision for one voice and one character, and both men left their families with an answer instead of a question. Executor Lock™ gives every adult the same power over their own face, voice and Persona, without a legal team. For the executor's role, read how Executor Lock™ works, and for the question of ownership, read why owning your AI identity matters.

The companies in this category have shown why the ending must be yours. HereAfter AI, which recorded parents' life stories for their families, went offline in August 2026, and families now have to email the company to ask for recordings, which is why our comparison of AI twin platforms reads what each contract says happens when you die. For how every Idy twin is governed after the owner dies, read how Timeless AI™ Twin passes to your executor.

Idy, the company behind Timeless AI™, settled who controls a person's face and voice before building any video of them. Most video-avatar companies started with rendering and left the questions of ownership, consent and death for later, and their contracts still leave those questions unanswered. We started with those questions, because a video of you is only as trustworthy as the rules behind the video, and our answers became Executor Lock™. The technology behind it is covered by Idy's 78 patents pending, which cover consent and permissions with proof that consent was given, verifying a death before control of a Persona changes hands, handing authority to a named executor, and sealing a Persona so the record cannot be altered after the handover.

That order of work is why Persona Video can do more than a video-avatar tool with the same footage. Your video comes from a recording you consented to, your voice and face live inside a Persona that knows your memories, and the people who take over when you die were named by you and accepted the Executor Lock™ Agreement while you were alive. Europe's governments have since written the same principle into a declaration signed under the Danish Council presidency by ministers from every EU member state except Hungary, stating that "Our citizens must be protected against digital replicas of their personal characteristics without consent" (European Parliamentary Research Service, January 2026). Persona Video was built on that rule from the start.

Record yourself, once, on a service that records your consent, keeps your videos with you and names who takes over when you die. Face a window, prop up your phone, give Persona Video 30 seconds of your face and your voice, and make your first message free. Then build the Persona behind the video, so your face and voice belong to an AI version of you that knows your stories, and read how to make an AI version of yourself for the full seven steps. If you want to compare every route to an AI clone of yourself, read how to clone yourself with AI.

1. Make your first video free

2. Name your executor

3. Put your twin online

We read the product, pricing and terms pages of HeyGen, Synthesia, Tavus and Creatify on October 1 and 2, 2026, the statutes, court rulings and legal commentary linked in the law table, and reporting from TechCrunch, PYMNTS, Popular Science, Fox 10 Phoenix, Fox News, Snopes, the World Economic Forum and the European Parliamentary Research Service. Every outside fact links to its source and date.

• how to make an AI version of yourself

• Timeless AI™ Twin, the twin that answers your clients as you

• Executor Lock™

• will your AI self survive model changes

• what an executor does in the first ninety days

• who owns your AI likeness after you die