A free online AI talking video generator tool
daVinci-MagiHuman is an open-source AI model that turns a single portrait plus text or audio into a lip-synced talking video, with audio and video generated in one pass. Self-hostable and released under Apache 2.0, it supports fast inference on NVIDIA GPUs.
Unified Audio + Video
Reference Photo Input
Multilingual
Open Source
- Upload a portrait image and enter your script or audio
- Choose resolution and run generation
- Self-host and use Hugging Face Hub for local/deployed setups
- Explore the GitHub repository for scripts and examples
Open-source with Apache 2.0 license
Single-pass generation for unified audio and video
Works from a single portrait photo
Fast inference on H100-class GPUs