Join our daily and weekly newsletters to get the latest updates and exclusive content on industry-leading AI coverage. Learn more
The competition between startups and large companies like Google and Meta to offer compelling AI video creation tools has entered a new era.
Luma Ia startup founded by a former Googler and others, is significantly expanding its Dream Machine AI video model with a new interface, mobile app and basic visualization model. Efficiency and creativity To push the limits of photo and video creation.
Now available on web and iOS, the new Dream Machine aims to combine simplicity and sophistication in input devices through a unified, intuitive conversational interface.
With more than 25 million registered users since its launch in June 2024, Dream Machine is evolving into a subscription-based service for both casual and professional creators in industries such as fashion, marketing, and filmmaking.
“We created Dream Machine as a visual thinking partner. It’s powered by a new imaging model called Luma Photon,” Amit Jain, CEO and co-founder of Luma AI, said in a video interview with VentureBeat. “It’s creative, intelligent, and designed for the people who shape our world. Whether it’s a designer Creator in fashion, media and entertainment.”
A new approach to image creation
Dream Machine aims to eliminate the complexity associated with traditional creative tools. Users can explain their ideas in natural language or provide reference images to guide the platform’s results. Unlike traditional engineering prompts that require precise technical information, Dream Machine is built for intuitive interactions.
“Unlike call engineering where you have to carefully craft specific commands, Dream Machine lets you talk to it like you’re talking to a person.” This conversational interface makes editing and creativity easy. Jain explained.
The platform’s new personalization features Includes multiple image prompts and single image character references. It allows users to realize their vision with greater precision and detail. For example, designers can upload textures, colors, and other images to guide the system’s output.
“With Dream Machine, you can give it a visual reference such as colour, structure or texture, and it will intelligently combine and iterate until you get exactly what you want. It is a game changer for designers and creators,” Jain added.
The new mode includes a brainstorming mode that allows users to apply different styles to their images and videos. as well as boards with multiple images and videos that can be shared between team members and fellow creators, and “Concept Pills” that offer presets. A unified stylistic image users can apply to their video and image output.
Corresponding characters from a single image
Luma AI imagines a future where creating images is as seamless and accessible as typing text. Dream Machine bridges this gap. Make advanced creation tools available to everyone From amateurs to industry professionals
“Why is it as difficult to create images and videos as it is to create images and videos? with using Adobe tools? Imagine if creating a message was that difficult. There would be no digital revolution. Visual thinking should be equally accessible,” Jain argued.
In addition to access The platform also offers innovative capabilities for video creation. Users can create animations with consistent characters drawn from a single image. This opens new doors for storytelling.
“Now you can create different styles. of a person endlessly from just one image This consistency results in all stories in the video featuring the same characters. This is something that has not been possible in video creation till now,” Jain said.
Photon reconstruction model
At the core of these advancements is the Luma Photon, the company’s latest image base model. which creates high quality still images from message notifications And it includes “state-of-the-art” embedded text, something that other image generation models do. Many are still experiencing problems. By achieving reliable results
Photon is built on Luma’s Universal Transformer architecture, which Luma says is up to 8 times faster and more cost-effective than comparable models.
This efficiency allows for rapid repeatability without compromising quality. This makes it ideal for demanding use cases.
“Our new Photon model is the most innovative and customizable model available today. It adjusts in real time without training. It uses an advanced universal transformation architecture,” explains Jain.
Developers can also leverage the power of Photon through the Luma AI API, which supports text-to-image conversion. text to video and image to video APIs ensure the privacy of user data and offer scalability for products built on the platform.
New membership price
The updated Dream Machine is available in four price levels:
- Hobbyists: $9.99/month
- Explorer: $29.99/month
- Expert: $99.99/month
- Enterprise: Custom pricing for large teams.
These levels offer flexibility to users with different needs. Whether they are built for personal projects or commercial ventures.
A new era of possibility
Since its founding in 2021, Luma AI has raised $80 million from strategic investors, including Andreessen Horowitz, Amplify Partners, Matrix Partners, General Catalyst, and South Park Commons.
Anjney Midha, general partner at Andreessen Horowitz, emphasizes the platform’s potential to move the industry. From creating virtual reality videos to creating interactive 3D worlds.
As Dream Machine evolves, Luma AI is delivering on its mission of democratizing creativity. “Dream Machine is where you come to visualize what’s in your head. It helps make sense of a more complex world by leveraging AI’s ability to process and simplify huge amounts of data,” Jain concludes.
With a combination of accessibility privacy and cutting-edge technology, Dream Machine is poised to redefine the way people create and share ideas in the digital age.
Correction: This article originally incorrectly stated the amount raised by Luma AI and incorrectly mentioned Amazon as an investor. This article has been updated to include the correct number and list of investors.
Source link
