OmniHuman 2.0
One photo. One voice. One video. — try your first one free, no signup.
Make any photo speak and perform
Upload a portrait and a voice — OmniHuman 2.0 turns them into a talking, performing video.
Portrait
Voice
First video is free — just your email, no credit card.
Example output — try it with your own photo & voice
The model, in plain words
What is OmniHuman 2.0?
OmniHuman 2.0 is an audio-driven digital human model released by ByteDance through BytePlus on October 15, 2026. Give it one portrait photo and one audio track, and it returns a video of that person speaking or performing. Speech, music, and dialogue all work as input.
Unlike classic lip-sync tools, the model reads rhythm and emotion in the audio. Head motion, facial expressions, and gestures follow the performance, so singing looks like singing. A single continuous take can run up to about five minutes.
This site is an independent third-party tool built on the model’s API. We render 30 to 300 second clips here, and your first video is free.
Photo in, video out
How it works
Model comparison
OmniHuman 2.0 vs 1.5
ByteDance shipped OmniHuman 2.0 through BytePlus on October 15, 2026. Here is how it compares with the previous generation, using publicly listed specs.
| Capability | OmniHuman 1.5 | OmniHuman 2.0 |
|---|---|---|
| Max single take | Up to 60 seconds at 720p, 30 seconds at 1080p | Up to about 5 minutes in one continuous take |
| Typical cost per second | Roughly $0.14–0.16 through third-party APIs | About $0.07 on the official API, roughly 40% lower |
| Visual consistency | Longer videos stitch segments, which can drift at cut points | One continuous take holds identity and motion; text prompts can steer performance |
Model specs and API rates are public listings as of October 2026 and may change. This site is an independent tool, not the model vendor.
Pay per video, not per month
Pricing
- First video is free — just an email, no credit card.
- After that, every 30-second video costs 30 credits.
- Failed renders refund their credits automatically.
- No subscription required. Buy credits only when you need them.
FAQ
Frequently asked questions
Third-party notice
omnihuman2.io is an independent tool and is not affiliated with, endorsed by, or sponsored by ByteDance or BytePlus. OmniHuman is a product name of ByteDance, and all trademarks belong to their respective owners. Model capabilities and pricing described on this page reflect public information as of October 2026.