In a significant update, Luma AI, a Uma-backed startup, has officially launched Ray3 Modify, a new generative video model that enables users to edit existing footage without losing the originality of content.
The recently launched app allows creators to offer character reference images, changing the actor’s appearance while maintaining the original emotional delivery, timings, and more.
Additionally, users can supply start and end frames to guide the generation of transitional footage, providing control over behaviour, scene transitions, and more.
This helps creative teams to edit footage without reshooting the scenes again and again, they can easily change locations, and customise it according to their own choice with advanced AI while preserving the original human performance.
“Generative video models are incredibly expressive but also hard to control,” said Amit Jain, co-founder and CEO of Luma AI. “Ray3 Modify blends real-world performance with AI expressivity while giving full control to creatives.”
Luma’s Ray3 Modify is currently accessible via Luma AI’s Dream Machine platform, after the company’s earlier video modification launch in June 2025.
With this significant update, Luma aims to compete with the leading players of the AI industry, including Runway and Kling.
The release follows a $900 million funding round led by Saudi Arabia’s Humain, alongside existing investors a16z, Amplify Partners, and Matrix Partners.
Alongside the update, Luma AI plans to build a 2GW AI cluster in Saudi Arabia with Humain to expand its generative AI capabilities.
