AI model · Video
OmniHuman 1.5
ByteDance’s OmniHuman 1.5, released in September 2025, creates a video from a single image of a person and an audio file.
- Provider
- ByteDance
- Type
- Video
- Cost
- From 177 credits
Overview
What is OmniHuman 1.5?
ByteDance’s OmniHuman 1.5, released in September 2025, creates a video from a single image of a person and an audio file. The character’s emotions, gestures and movements closely follow the audio, well beyond the lips.
In Havincy, upload a portrait (photo or generated image) and a voice: OmniHuman 1.5 makes it talk or sing, in 720p or 1080p. The length follows the audio. It is the most expressive talking-avatar model in the catalog.
Key strengths
- Talking avatar
- Audio→vidéo
- Corps+lèvres
Specs
OmniHuman 1.5: specifications
- Publisher
- ByteDance
- Generation modes
- talking avatar (image + voice)
- Resolutions
- 720p, 1080p
- Available since
- September 23, 2025
- License
- Commercial use allowed
- Cost in Havincy
- From 177 credits
Specs taken from the publisher’s public model page. See the documentation on fal.ai
Use cases
What you can create with OmniHuman 1.5
Brand avatar
A virtual spokesperson presenting your products from a single photo.
Training and e-learning
Educational videos with an animated trainer, no shoot required.
Singing character
Make a character sing to generated music.
UGC content
Realistic to-camera videos for your ads.
Prompt examples
Prompts to try with OmniHuman 1.5
Paste an example into Havincy, then adapt the subject, style and format to your project.
Make this portrait speak with the voiceover I just generated, enthusiastic tone, natural gestures.
Animate this photo of our founder so she presents the new collection.
Make this character sing this 15-second chorus.
Use cases
Compatible skills
Use this model with the following Havincy skills.
How it works
How to create with this model
- Describe the desired result in a Havincy conversation.
- Add a reference when the style, product, or character needs to be preserved.
- Generate, then ask for a variation or improvement without starting over.
Examples made with this model
FAQ
Frequently asked questions about OmniHuman 1.5
How much does OmniHuman 1.5 cost?
On fal, $0.16 per second of video. In Havincy, about 177 credits for 8 seconds.
OmniHuman 1.5 or OmniHuman?
Version 1.5 improves expressiveness and the link between emotions, gestures and audio, and lets you pick 720p or 1080p. OmniHuman stays slightly cheaper.
What do I need?
An image of a person (face clearly visible) and audio. You can generate both in Havincy: the portrait with an image model, the voice with ElevenLabs or MiniMax.
Is OmniHuman 1.5 open source?
No, it’s a ByteDance model offered through the fal API. In Havincy, you use it directly with your credits.
More models to discover
A starting point for your next creation.
Ready to create with OmniHuman 1.5?
Describe your idea in Havincy, then refine the result in the conversation.