HeyGen Avatar 6
Personal Avatars that
show up as your best self
Our smartest, most powerful avatar model yet. It understands your script and performs it the way you want, on the first take.

Brynn · Real Avatar 6 output
Presenter
Your avatarAdd a photo + short clip after sign-up
Sign up free, make your avatar from a photo and a short clip, and your first video says this.
Built for the videos you’d film yourself
One real person, any look, any place. Avatar 6 gets the feeling and the gestures from the script the first time.

0:00 / 0:00
Core features
What’s new in Avatar 6
Previous avatars looked like you but recycled the same gestures. Avatar 6 reads the script and performs it the way you would.

01
Higher realism
Lip sync, motion, and realism all improved. Every frame is native 2K, and side by side with real footage, Avatar 6 is 84.4% indistinguishable, up from 44.4% for Avatar V.

02
Context-aware gestures
Avatar 6 understands what is being said and gestures to match, including pointing, counting, and directional motion. For finer control, prompt the emotion, expressions, and gestures yourself.

03
Fresh motion, every line
Rich, varied movement that looks and moves like you, matched to your script rather than replayed from a loop.

04
Great the first time
No tweaking, no regenerating, no settling. The video you generate is the video you love.

05
Two avatars, one conversation
Put two avatars in the same frame and let them talk. Write both sides of the script and get a full host-and-guest conversation, generated in one go.
Try two avatars in MCP
The research
Tested against the
strongest video models
We put Avatar 6 up against Seedance 2.5, Wan 3.0, our previous models, and commercial avatar products. It was judged by people in the videos, external raters, and side-by-side comparisons with real footage.
| Single clip, shown alone | Judged real | Answers |
|---|---|---|
| Real recording | 78.9% | 612 |
| Avatar 6, person’s own voice | 65.1% | 304 |
| Avatar 6, HeyGen Voice | 56.5% | 308 |
| Side by side, same sentence and audio | Real picked | Answers |
|---|---|---|
| Real vs Avatar 6 | 57.8% | 270 |
| Avatar 6 picked as real | 42.2% | 270 |
| Indistinguishability (2× the above) | 84.4% | 270 |
From a single photo and speech
- Avatar 6Overall 4.00
- VEED Fabric 1.0Overall 3.68Avatar 6 wins 69.6%
- Argil image avatarOverall 3.46Avatar 6 wins 80.9%
- OmniHuman v2Overall 3.32Avatar 6 wins 80.8%
- Creatify AuroraOverall 3.29Avatar 6 wins 84.3%
- Akool Talking PhotoOverall 3.19Avatar 6 wins 86.4%
- Hedra Character 3Overall 3.16Avatar 6 wins 88.2%
- Colossyan NEO 1 Instant avatarOverall 1.59Avatar 6 wins 98.1%
- D-ID V2 TalksOverall 1.48Avatar 6 wins 97.9%
| Method | Overall | Visual quality | Identity | Lip sync | Motion naturalness | Gesture–speech | Artifacts, stability | Avatar 6 wins |
|---|---|---|---|---|---|---|---|---|
| From a single photo and speech | ||||||||
| VEED Fabric 1.0 | 3.68 | 3.96 | 4.20 | 3.71 | 3.74 | 3.95 | 3.70 | 69.6% |
| OmniHuman v2 | 3.32 | 3.70 | 3.97 | 3.44 | 3.41 | 3.48 | 3.36 | 80.8% |
| Argil image avatar | 3.46 | 3.76 | 4.32 | 3.52 | 3.55 | 3.64 | 3.58 | 80.9% |
| Creatify Aurora | 3.29 | 3.33 | 3.99 | 3.48 | 3.40 | 3.58 | 3.32 | 84.3% |
| Akool Talking Photo | 3.19 | 3.52 | 4.11 | 3.28 | 3.33 | 3.35 | 3.40 | 86.4% |
| Hedra Character 3 | 3.16 | 3.62 | 4.05 | 3.24 | 3.21 | 3.28 | 3.33 | 88.2% |
| D-ID V2 Talks | 1.48 | 3.43 | 3.07 | 2.18 | 0.94 | 0.71 | 2.42 | 97.9% |
| Colossyan NEO 1 Instant avatar | 1.59 | 3.40 | 3.48 | 1.76 | 1.19 | 0.78 | 2.57 | 98.1% |
| Avatar 6 | 4.00 | 4.33 | 4.39 | 4.12 | 4.01 | 4.07 | 4.08 | – |
| From the person’s own footage | ||||||||
| Akool Talking Avatar | 2.93 | 3.70 | 3.89 | 2.78 | 2.85 | 2.72 | 3.29 | 92.9% |
| D-ID V3 Express avatar | 2.85 | 3.74 | 3.83 | 2.60 | 2.76 | 2.71 | 3.23 | 95.3% |
| Colossyan NEO 2 / NEO 1 | 2.68 | 3.59 | 3.65 | 2.56 | 2.69 | 2.48 | 3.21 | 97.3% |
| NEO 2, scripts 1–3 | 2.90 | 3.59 | 3.58 | 3.24 | 2.86 | 2.58 | 3.35 | 96.2% |
| NEO 1, scripts 4–5 | 2.35 | 3.60 | 3.74 | 1.55 | 2.43 | 2.34 | 3.01 | 98.9% |
| Avatar 6 | 3.98 | 4.24 | 4.19 | 4.10 | 4.04 | 4.08 | 3.93 | – |
Footage input
- Kling 4.0 reference-to-videoAvatar 6 wins 54.9%Overall lead +0.17
- Seedance 2.5 reference-to-videoAvatar 6 wins 72.4%Overall lead +0.61
- Wan 3.0 reference-to-videoAvatar 6 wins 74.1%Overall lead +0.75
- VEED Preset avatarsAvatar 6 wins 87.6%Overall lead +1.15
- Creatify AI Avatar v1Avatar 6 wins 93.3%Overall lead +1.31
- Synthesia Personal video avatarAvatar 6 wins 93.6%Overall lead +1.68
- Synthesia Stock avatars v3 / v1Avatar 6 wins 93.7%Overall lead +1.28
- v3, four avatarsAvatar 6 wins 92.8%Overall lead +1.29
- v1, CooperAvatar 6 wins 97.5%Overall lead +1.26
Can you tell which one is real?
One clip is a real recording. The other is Avatar 6. Pick the one you think is real.
Showcase
Made with Avatar 6
Real creators, real scripts, zero cameras. Every clip was generated with Avatar 6, and most of them from a single photo.
How it works
How to use Avatar 6
A few simple steps take you from first recording to finished video.

Record your avatar
Upload a photo, verify it’s you, then record at least 15 seconds of yourself talking. Avatar 6 picks up your voice, expressions, and motion.

Write your script
Bring your own script, upload audio, or just record yourself talking. Starting from scratch? Drop in a prompt and HeyGen writes the script for you.

Generate your video
Set your voice, presenter style, and pace, then hit generate. Now your finished video is ready to download and share.
The world’s best avatar model
See it for yourself. Your first video is on us, no credit card needed.

FAQs
What is HeyGen Avatar 6?
It’s our newest avatar model. Give it one photo and it makes a video of you that acts out your script like you would. Side by side with the real recording, raters pick Avatar 6 as the real one 42.2% of the time. 50% would mean they cannot tell at all, so Avatar 6 is 84.4% of the way there.
How is it different from Avatar V?
Avatar V looked like you but reused the same gestures. Avatar 6 reads your script and gestures to match, like pointing or counting on fingers, and the lip sync looks more real.
Is it free?
Your first video is free, and you don’t need a card. After that it uses credits, with plans starting at $29/month for 600 credits.
How do I make one from a photo, and how long does it take?
Upload a clear photo of your face. For an avatar of yourself, also record 30 seconds of video. Then add your script and hit generate.
What resolution does it export?
Every frame is made in 2K. You can export in 1080p or 2K.
Can I use it through the API or MCP?
Yes, both. Developers can use the API, or you can just ask Claude or ChatGPT to make a video through HeyGen MCP. It can even put two avatars in one conversation.
Who can make an avatar of me, and how is my likeness protected?
Only you, or someone you’ve given clear permission to. Making an avatar of someone without their consent is against our rules. You can ask us to remove your likeness anytime at [email protected].



