Power video generation with digital twins
Help your users scale video creation 
with our easy-to-use APIs
A better way to create videos
Users understand the power of video to engage audiences. But recording yourself at scale is impossible.
With just a few simple API calls, you can implement video generation using realistic digital twins into your platform.
With just a few simple API calls, you can implement video generation using realistic digital twins into your platform.
Video generated by tavus
The training footage used to generate the personal replica
Why product development teams build video generation with Tavus
It's just a couple of API calls
Train replicas with one API call, and generate video with another API call.
Training is simple
Users only need to submit about two minutes of video footage to train their digital replica.
Photo-realistic replicas
Our in-house models deliver lifelike replicas. We blend new techniques including audio and text-driven models like 3D Gaussian Splatting.
Generate in 30+ languages
Users can submit scripts in over 30 languages and easily generate foreign language videos using their real voice.
Bring your own audio
Use our default TTS or bring your own audio files. For example, users can record natural audio to generate videos.
Get into the details of our industry leading developer platform
See Developer Docs
Safe usage is handled for you
Consent Management
A person is only able to create a replica of themselves. We employ a suite of safety checks including voice identification and user consent to ensure only a user can make a replica of themselves, and hold the keys to their likeness.
Content Moderation
All scripts submitted by users undergo review to ensure they pass content safety standards. Suspicious content is flagged and blocked if deemed unsafe. Enterprise customers can customize content guidelines to meet specific needs.
Leverage powerful APIs to accelerate time to market
We’ve done the hard work for you.
Create Replica Endpoint
Creates a new replica that can be used to generate videos.
A user submits a 2 minute training video. The only required body parameter is train_video_url.
Replica training requires a consent statement from the user.
Create Video Endpoint
Generates a new video using a replica and a written script or an audio file.
It’s possible to include a background like a recording of a website or a video/image file.
If no background is provided, the video will consist of a full screen Replica.
Start building with dead simple APIs.
Get started today.