Luma AI has released an update for its AI video generator Dream Machine. Version 1.6 now generates videos ten times faster.
Luma AI has released the latest version of its Dream Machine AI video generator. With version 1.6, users can now generate full-quality videos in less than 20 seconds - a tenfold acceleration compared to the previous version 1.5. The company emphases that this increase in speed has been achieved without compromising on quality.
Video: Luma AI
Another major innovation in Dream Machine 1.6 was the introduction of camera moves at the beginning of September. Users can now simply add the word "camera" to their prompts to add different camera movements to their generated videos.
Video: Luma AI
Dream Machine is based on a scalable, efficient and multimodal transformer architecture that is trained directly on video. According to Luma AI, this enables the generation of physically accurate and consistent scenes. The model is said to have an understanding of how people, animals, and objects interact with the physical world.
Luma AI opens Dream Machine to developers
Luma AI also recently released an API for Dream Machine that allows developers and businesses to integrate AI video generation into their own applications. The API costs $0.32 per million pixels generated, or approximately $0.35 for a five-second video at 720p resolution.
The Dream Machine API offers features such as text-to-video generation, image-to-video conversion and camera motion control.
With the launch of the API and recent enhancements, Luma AI is positioning itself in an increasingly competitive AI video generation market that includes companies such as Runway, Adobe and OpenAI.
