Turn one portrait image and an audio track into a natural talking-person video with OmniHuman 1.5. Create digital presenters, localized messages, lessons, character performances, and social clips with synchronized lips, facial expression, and body motion.
Create a video with a focused brief and the source media required by the selected workflow.
OmniHuman 1.5 Guide
A practical guide to the model's inputs, core capabilities, ideal use cases, and prompt strategy.
OmniHuman 1.5 is a ByteDance human-animation model that turns a single image and audio track into a talking-person video with synchronized lips, expression, and motion.
Turn one portrait image and an audio track into a natural talking-person video with OmniHuman 1.5. Create digital presenters, localized messages, lessons, character performances, and social clips with synchronized lips, facial expression, and body motion.
Animate a single portrait with a supplied voice track, removing the need for source video footage. The image defines appearance while audio drives the timing and performance. Generate mouth motion, facial expression, and subtle head behavior that follow the rhythm and emotion of speech for a more convincing talking-person result.
OmniHuman 1.5 extends beyond a static talking head by producing coordinated posture and gesture where the portrait composition provides enough visible body context. Create digital presenters, educational explainers, character dialogue, personalized messages, and multilingual campaigns from reusable visual identity and new audio.
Upload a clear image with a visible face and enough space around the head or upper body for natural movement. Use clean voice audio with minimal background noise and intentional pacing. The audio controls the performance timing.
Create the clip, review lip sync and motion from start to finish, then download or adjust the source image or audio.
Primary references
Animate a single portrait with a supplied voice track, removing the need for source video footage. The image defines appearance while audio drives the timing and performance.
Try OmniHuman 1.5Generate mouth motion, facial expression, and subtle head behavior that follow the rhythm and emotion of speech for a more convincing talking-person result.
Try OmniHuman 1.5OmniHuman 1.5 extends beyond a static talking head by producing coordinated posture and gesture where the portrait composition provides enough visible body context.
Try OmniHuman 1.5Create digital presenters, educational explainers, character dialogue, personalized messages, and multilingual campaigns from reusable visual identity and new audio.
Try OmniHuman 1.5Create a video with a focused brief and the source media required by the selected workflow.
Upload a clear image with a visible face and enough space around the head or upper body for natural movement.
Use clean voice audio with minimal background noise and intentional pacing. The audio controls the performance timing.
Create the clip, review lip sync and motion from start to finish, then download or adjust the source image or audio.
OmniHuman 1.5 is a ByteDance human-animation model that turns a single image and audio track into a talking-person video with synchronized lips, expression, and motion.
Animate one image with your audio to create expressive presenters and character performances.
Try OmniHuman 1.5