Luma AI launches Ray3

The world’s first reasoning video model and the first to generate high fidelity 16 bit HDR.

Ray3 introduces breakthrough new reasoning capabilities in the world's first multimodal reasoning system, native HDR EXR generations, and Draft Mode for faster creative exploration

Adobe is the first partner to launch Ray3 outside of Dream Machine, helping drive global adoption at scale with the Adobe Firefly app

Collaborating with Dentsu Digital to revolutionize video advertising production in Japan with Ray3

Additional global agency launch partners include HUMAIN Create, Monks (S4), Galeria, and Strawberry Frog, bringing Ray3 to film, advertising, and broader creative industries worldwide

PALO ALTO, Calif., September 18, 2025 – Luma AI, a frontier generative AI lab, today unveiled Ray3, the first video model built to think like a creative partner. Available now in the Dream Machine platform, Ray3 empowers filmmakers, advertisers, and game developers to move from idea to final cinematic video faster than ever before, while meeting the technical standards of professional production.

Unlike earlier models that could only generate visuals with limited control, Ray3 is the first system that can reason in visuals and concepts, evaluate its own outputs, and refine its results on the fly. It’s also the first model to produce video in true 10-, 12-, and 16-bit High Dynamic Range (HDR) ACES2065-1 EXR format, making it suitable for high-end film and advertising pipelines. And with its new Draft Mode, creatives can explore dozens of ideas up to 20x faster, then select and polish the best options into cinematic 4k quality HDR.

“Ray3 is our first step toward building intelligence for creative work,” said Amit Jain, CEO and co-founder of Luma AI. “Creative work is one of the most intellectually challenging things humans do, yet until now, many of the AI available to creatives has lagged far behind what’s possible in coding and analysis with language models. Generative models have been more like slot machines - powerful but not intelligent. Ray3 changes that in a big way. Its groundbreaking reasoning system can understand intent, evaluate its own outputs, and refine results, significantly improving the accuracy and quality of generated video. More than twice the size of Ray2, Ray3 delivers new levels of fidelity, instruction following, and temporal coherence — and it’s the first model capable of generating High Dynamic Range video in ACES EXR, making it purpose-fit for high-end production pipelines. This is the leap forward that filmmakers, advertisers, and game developers have been waiting for.”

Luma AI is introducing Ray3 alongside some of the most influential names in technology, advertising, and creative services, underscoring the model’s broad global potential.

Adobe and Luma AI are partnering to bring Ray3 directly into the Adobe Firefly app, Adobe’s all-in-one destination for AI-powered ideation, creation and production. With this integration, Adobe becomes the first partner to launch Ray3 outside of Luma AI’s Dream Machine platform. Starting today, Ray3 will be available globally in the Firefly Video module and Firefly Boards. For customers who want more precise editing and refinement, Ray3 generations in Firefly can be brought into Adobe’s professional-grade tools like Premiere Pro. To help users get started, Adobe is offering unlimited Ray3 generations for the first 14 days to customers on a paid Firefly or Creative Cloud Pro plan. Read more about Ray3 in the Firefly app here.

“Adobe is building the creative AI ecosystem of the future with Adobe Firefly - our all-in-one destination to access the industry’s top creative AI models, including our commercially safe Firefly models and choice of models from trusted partners like Luma AI – all integrated into the tools creators know and use every day,” said Hannah Elsakr, Vice President, New GenAI Business Ventures, Adobe. “With Ray3 now available in the Firefly app, Adobe customers are among the first to gain access to a powerful new video model that amplifies imagination and transforms workflows. We can’t wait to see how they use it to bring their ideas to life.”

Dentsu Digital, one of the largest integrated digital firms in Japan, is collaborating with Luma AI to bring Ray3-powered, AI-accelerated advertising to Japan. As the sole launch partner in the market, Dentsu Digital will integrate Ray3 into its production pipelines to give global and domestic brands new levels of personalization, agility, and storytelling power.

"Dentsu Digital has always sought to push the boundaries of how brands connect with audiences, and Ray3 allows us to do this faster and with more impact than ever before," said Satoru Yamamoto, CAIO (Chief AI Officer) and Executive Officer of Dentsu Digital. "Together with Luma AI, we are positioned to lead the Japanese market in applying AI-driven storytelling to advertising, delivering more dynamic, relevant, and effective content for our clients."

In the MENA region, HUMAIN Create is bringing Ray3 to both creators and enterprises, delivering AI content generation grounded in Middle Eastern and Islamic culture, values, history, and heritage, and expanding access across one of the fastest-growing regions for digital content.

“Ray3 isn’t just an upgrade, it’s a quantum leap. By giving AI the power to reason across words, images, and motion, we’re not only supercharging the speed and fidelity of creative output, but we’re also weaving in smarter guardrails. That means sharper ideas delivered faster, and safer content that respects ethics, compliance and cultural context. It’s a rare moment where creativity, trust, and technology all level up together,” said Steve Plimsoll, Chief Strategy Officer of HUMAIN.

With creative leaders like Monks, Galeria, and Strawberry Frog also adopting Ray3, Luma is partnering with iconic creatives known for shaping culture-defining campaigns and pushing the boundaries of design, film, and brand storytelling. Together, these partnerships reflect not only market confidence in Ray3 but also its immediate impact across industries where creativity meets scale.

Breakthrough Features of Ray3

• Multimodal Reasoning System – At the heart of Ray3 is a new type of intelligence designed specifically for creative work. Instead of just responding randomly to prompts, Ray3 can actually think through what a user is asking, plan complex scenes, and judge whether its own output makes sense. It does this by generating both text and visual tokens, almost like a director sketching out a storyboard before filming. This means creatives can give more precise instructions, such as adding annotations directly to images, and the model can follow through with complex, multi-step ideation. The result is videos that feel more coherent, with characters that look consistent, scenes that unfold naturally over time, and physics that behave as they should.
• Native HDR Generation – For the first time in generative AI, Ray3 creates video in true High Dynamic Range (HDR) using the professional ACES2065-1 EXR standard across 10-, 12-, and 16-bit formats. In practical terms, this gives filmmakers and advertisers access to the same deep color and exposure controls they would expect from footage shot on high-end cameras. Creators can even take regular SDR videos, whether generated in Dream Machine or captured with a camera, and convert them into generative HDR for richer color and more flexibility in editing. This makes Ray3’s output ready not just for experimentation, but for real production pipelines.
• Draft Mode for Iteration – One of the biggest bottlenecks in creative work is the time it takes to try out ideas. Ray3 solves this with Draft Mode, a new modethat produces test videos up to 10 times faster than usual. Artists can generate a wide range of options quickly, then choose the ones they like best to “master” into high-resolution, full-quality outputs. Importantly, Draft Mode preserves the identity, motion, and composition of the draft when moving to the final render, making sure the creative vision isn’t lost in transition. This lets creators enter a state of flow, experimenting freely without worrying about time or compute costs.
• Native 1080p Generations – Ray3’s architecture has been scaled up to produce crisp, detailed 1080p video natively, a major leap in resolution over past models. For projects that demand even more polish, Ray3 includes a neural upscaler that cleanly upscales the output to 4K without introducing blur or motion artifacts. For professionals in advertising, film, and gaming, this means AI-generated video can finally meet the technical requirements of broadcast and studio production. At launch, native 1080p is available to select partners in early access and will become generally available soon.
• Powerful Controls – Ray3 makes the tools creators already know and love much more powerful and expressive. Image-to-Video lets users animate still images; Keyframes provide more control over timing and scene changes; Extend allows users to grow a shot beyond its original length; and Loop enables seamless, repeating animations. In Ray3, each of these controls has been made smarter and more precise, opening up new styles and workflows for a wide range of creative projects.

About Luma AI

Luma AI is building a multimodal general intelligence that can generate, understand, and operate in the physical world. Its flagship platform, Dream Machine, enables creatives everywhere to generate professional-grade video and images, and now powers creation for over 30 million users. In 2025, Luma released Ray2, a frontier video generative model capable of creating realistic visuals with stunning detail and natural motion. Luma’s models are utilized by top entertainment studios, ad agencies, and technology leaders, including Adobe and AWS, and are available via subscription or API. The company is backed by Andreessen Horowitz, AWS, AMD, NVIDIA, Amplify Partners, Matrix Partners, and angels from across technology and entertainment space.

Contact

Peter Binazeski
Head of Communications
Luma AI
peter@lumalabs.ai