Luma AI has officially introduced Ray3, describing it as the first generative AI model able to produce professional-grade HDR videos.
For many in Hollywood, this might be the announcement they were waiting for. Ray3, according to its creators, represents a breakthrough: a video model capable of delivering 16-bit color depth, EXR export for streamlined integration into professional workflows, and support for SDR conversion.
What makes Ray3 different
One of Ray3’s biggest strengths, according to Luma AI, lies in its reasoning abilities. Unlike traditional tools, it can accurately follow complex instructions, providing greater control over objects, movements, and camera angles.
Luma describes it this way:
Ray3 is an intelligent video model designed for storytelling. It can think and reason visually, bringing cutting-edge physics and coherence. For the first time, it generates videos with 16-bit High Dynamic Range colors, making generative video production suitable for professional studio pipelines.
Beyond text prompts: direct scene interaction
Another key innovation is the ability to edit scenes visually, by directly interacting with individual elements, instead of relying only on text prompts (which AI doesn’t always interpret correctly).
The only major limitation right now is the lack of automatic audio generation, something already available in Google’s Veo 3. Still, Ray3 can be tested for free on Dream Machine, and demo videos showcasing its potential are already available on the official website.
AI and cinema: a growing connection
The film industry is watching AI developments with increasing interest. Heavyweights like James Cameron have already spoken out, showing direct involvement in the debate. Meanwhile, OpenAI is working on Critterz, an animated film expected to hit theaters by 2026.
