Luma AI launches Ray3 Modify to generate video from start and end frames with character consistency

Luma releases a new AI model that lets users generate a video from a start and end frame | TechCrunch

Luma, the a16z-backed AI video and 3D model company, released a new model called Ray3 Modify that allows users to modify existing footage by providing character reference images that preserve the performance of the original footage. Users can also provide a start and an end frame to guide the model to generate transitional footage. The company said Thursday the Ray3 Modify model solves the problems of preserving human performance while editing or generating effects using AI for creative studios. The startup said the model follows the input footage better, allowing studios to use human actors for creative or brand footage. Luma mentioned the new model retains the actor's original motion, timing, eye line, and emotional delivery while transforming the scene. With Ray3 Modify, users can provide a character reference for transformation to the original footage and convert the human actor's appearance into that character. This reference also allows creators to retain information like costumes, likeness, and identity across the shoot. What's more, users can also provide start and end reference frames to create a video using the new Ray3 Modify model. This is helpful for creators in directing transitions or controlling character movements or behaviour while maintaining continuity between scenes. "Generative video models are incredibly expressive but also hard to control. Today, we are excited to introduce Ray3 Modify that blends the real-world with the expressivity of AI while giving full control to creatives. This means creative teams can capture performances with a camera and then immediately modify them to be in any location imaginable, change costumes, or even go back and reshoot the scene with AI, without recreating the physical shoot," Amit Jain, co-founder and CEO of Luma AI, said in a statement. Luma said the new model is available to users through the company's Dream Machine platform. The company, which competes with the likes of Runway and Kling, released video modification capabilities in June 2025. The model release comes on the back of a fresh $900 million funding round, led by Saudi Arabia's Public Investment Fund-owned AI company Humain, for the startup announced in November. Existing investors like a16z, Amplify Partners, and Matrix Partners also participated in the round. The startup is also planning to build a 2GW AI cluster in Saudi Arabia along with Humain.

Luma AI brings character consistency to video with Ray3

Luma AI, an a16z-backed company specializing in AI video and 3D models, released Ray3 Modify, a new model that enables users to modify existing footage using character reference images while preserving the original performance. Users supply start and end frames to generate transitional footage, addressing preservation challenges for creative studios. Ray3 Modify tackles issues in maintaining human performance during AI-based editing or effects generation. The model adheres closely to input footage, permitting studios to employ human actors in creative or brand productions. It specifically retains the actor's original motion, timing, eye line, and emotional delivery as it transforms the scene. Users input a character reference to alter the human actor's appearance into the specified character within the footage. This process preserves details such as costumes, likeness, and identity throughout the entire shoot. Such capabilities ensure consistent visual elements across sequences without additional filming. Creators also supply start and end reference frames to produce video segments with Ray3 Modify. This feature assists in directing smooth transitions or specifying character movements and behaviors. Continuity between scenes remains intact, supporting complex narrative structures in video projects. Amit Jain, co-founder and CEO of Luma AI, stated, "Generative video models are incredibly expressive but also hard to control. Today, we are excited to introduce Ray3 Modify that blends the real-world with the expressivity of AI while giving full control to creatives. This means creative teams can capture performances with a camera and then immediately modify them to be in any location imaginable, change costumes, or even go back and reshoot the scene with AI, without recreating the physical shoot." The model integrates directly into Luma's Dream Machine platform, making it accessible to users immediately. Luma positions itself against competitors including Runway and Kling in the AI video generation space. The company introduced initial video modification capabilities back in June 2025. This release follows a $900 million funding round announced in November, led by Humain, an AI company owned by Saudi Arabia's Public Investment Fund. Existing investors a16z, Amplify Partners, and Matrix Partners joined the round, providing substantial capital for expansion. Video: Luma Luma plans to construct a 2GW AI cluster in Saudi Arabia in collaboration with Humain. This infrastructure initiative supports scaling computational resources for advanced AI video and 3D modeling operations.