AAIHow

Make a video with your own cloned voice

You'll end up with: a 60-second AI video using your real voice

Overview
30–45 min
Beginner
Free to start
3 tools
Cost breakdown
Claude (script)Free
ElevenLabs (voice)Free tier
HeyGen (video)1 free credit
Captions.aiFree
TotalFree to start
Common mistake

Recording your voice in a noisy room. Even a small amount of echo makes the clone sound robotic. Record in a walk-in closet or hang blankets around you. Takes 2 minutes to set up and makes a massive difference.

Related guides
Clip long videos into social shortsTurn a blog post into a videoWrite a blog post in minutesCreate social media graphics
Before you start
  • A quiet room and a microphone — your phone mic works fine
  • A written script, 60–90 seconds when read aloud (roughly 150–220 words)
  • Free accounts on ElevenLabs and HeyGen — takes 3 minutes to sign up
  • No coding, no technical experience needed
1

Write your script with AI

Use Claude to write a natural, spoken-word script. Short sentences work best — they sound better out loud.

Claude.aiFreeOpen Claude.ai
Exact action

Write a 90-second spoken script about [your topic]. Use short sentences. Write how a real person talks — no bullet points, no headers, no formal language. Each sentence should flow naturally into the next. End with one clear call to action.

Good output: reads like someone talking, not writing. You should be able to read it aloud without sounding robotic.
If it sounds stiff or uses phrases like "In conclusion" — paste it back and say: "Make this sound more casual and conversational."
2

Clone your voice

Record yourself reading anything for 1–2 minutes. ElevenLabs uses this to build your voice model.

ElevenLabsFree tierOpen ElevenLabs
Exact action

1. Go to elevenlabs.io → sign up free 2. Click "Voices" in the left sidebar → "Add Voice" → "Instant Voice Clone" 3. Record yourself reading your script (or any text) for 1–2 minutes 4. Name your voice — e.g. "My Voice" 5. Click "Add Voice" and wait 30 seconds 6. Now paste your script into the text box and click Generate 7. Download the MP3

Good result: sounds like you, maybe slightly smoother. The AI cleans up background noise and hesitations.
Sounds robotic or wrong? Your recording had too much background noise. Record in a quieter spot — a closet with clothes works great.
3

Generate your AI video

Upload your cloned voice audio to HeyGen. It syncs your voice to an AI avatar that lip-syncs perfectly.

HeyGen$29/mo or 1 free creditOpen HeyGen
Exact action

1. Go to heygen.com → sign up (you get 1 free video credit) 2. Click "Create Video" → "Avatar Video" 3. Pick an avatar — or upload a photo of yourself 4. Under "Audio" → select "Upload Audio" → upload your ElevenLabs MP3 5. Paste your script into the script box (for captions) 6. Click Generate — takes 2–4 minutes 7. Download your video

Good result: the avatar's lips match your audio naturally. Slight imperfections are normal and not noticeable to most viewers.
Lips out of sync? This happens if your audio has long silences. Go back to ElevenLabs and regenerate without pauses, or trim silence in a free tool like Audacity.
4

Add captions and publish

Captions increase views by 40% on average. Captions.ai does this in under 2 minutes for free.

Captions.aiFreeOpen Captions.ai
Exact action

1. Go to captions.ai → sign up free 2. Upload your HeyGen video 3. It auto-generates captions — review and fix any errors 4. Choose a caption style (bold + centered works best for social) 5. Export and download 6. Upload to LinkedIn, Instagram, TikTok, or wherever your audience is

Done. You have a professional AI video with your voice, an avatar, and captions — in under an hour, for free.
Caption timing off? Drag the caption blocks in the editor to align them. Takes 2 minutes to fix manually.

All done!

You now have: a 60-second AI video using your real voice

Explore more guides