Luma AI has officially launched Ray2, its latest generative video model, which is now accessible through the Dream Machine platform for paid subscribers. This release marks a significant advancement in AI-driven video creation, offering users the ability to generate high-quality, realistic videos from text prompts in seconds. Initially focused on text-to-video capabilities, future updates will include image-to-video, video-to-video, and editing functionalities.
Introducing Ray2, a new frontier in video generative models. Scaled to 10x compute, #Ray2 creates realistic videos with natural and coherent motion, unlocking new freedoms of creative expression and visual storytelling. Available now. Learn more https://t.co/jGI6KmRQpR. pic.twitter.com/i9PYlatlPv
— Luma AI (@LumaLabsAI) January 15, 2025
Key Features of Ray2
Ray2 leverages a multimodal transformer architecture trained directly on video data, enabling it to produce visually coherent and dynamic 5-10 second clips. The model excels in:
- Natural Motion: It creates lifelike animations with smooth transitions and advanced cinematography.
- Instruction Following: Users can provide detailed text prompts to achieve precise visual outputs, such as a "man running through snow with explosions" or "hands slicing steak with rising steam."
- Physics Accuracy: Improved understanding of real-world interactions between objects, animals, and humans enhances realism.
- Resolution and Duration: Supports up to 1080p resolution and is reportedly being tested for extended video durations of up to 20 seconds.

Availability and Access
Ray2 is currently available to paid subscribers of Dream Machine. Luma AI also plans to integrate the model into Amazon's AWS Bedrock platform, broadening its reach to developers and businesses. The company aims to cater to a wide audience, including creators, marketers, and professionals in industries like gaming, entertainment, and e-commerce.
Comparison with Competitors
Ray2 positions itself as a leader in the generative video space by offering faster rendering times (under 10 seconds) and higher-quality outputs compared to models like Runway's Gen-3 or OpenAI's Sora. Its focus on advanced cinematography and real-world physics sets it apart from earlier models that struggled with motion accuracy or visual consistency.
Future Implications
The release of Ray2 underscores Luma AI's commitment to advancing generative AI technologies. By simplifying the creation of professional-grade videos without extensive resources or expertise, Ray2 has the potential to revolutionize content creation across various sectors. As additional features roll out, the model could further solidify its position as a go-to tool for creative professionals.