Luma AI made waves with the launch of its AI Dream Machine generative video creation platform last summer.
Of course, although this was only seven months ago, the AI video space has expanded rapidly with the release of many recent AI video creation models from competing startups in the US and China, including Runway, Kling, Pika 2.0, Sora OpenAI, Google Veo 2, Hailuo by MiniMax, and open source alternatives akin to Hotshot and Mochi 1 by Genmo, to call a few. Even Luma itself recently updated its Dream Machine platform, adding recent whiteboards for generating still images and brainstorming, and also debuted an iOS app.
But the updates proceed: San Francisco-based Luma launched today Ray2, its latest AI video generation model, available now via the Dream Machine website and mobile apps for paying subscribers (for starters).
According to the co-founder and CEO of Luma AI, the model offers “fast, natural, consistent movement and physics.” Amit Jain on his X accountand was trained with 10 times more processing power than the original Luma AI video model, Ray1.
“This dramatically increases the success rate of usable, production-ready generations and makes video storytelling accessible to many more people,” Jain added.
Luma Dream Machine online platform offers a free tier of 720 pixel generations with a variable limit each month: paid plans start at $6.99 per thirty days: from “Lite” offering 1080p graphics, through Plus ($20.99/month), to Unlimited (66 .49/month) and Enterprise ($1,672.92/yr).
Leap forward in the video gene
Right now, Luma’s Ray2 is limited to text-to-video, allowing users to type descriptions that turn into 5-10 second video clips.
The model can generate recent videos in seconds, although currently it might probably take several minutes as a consequence of high demand from recent users.
Examples provided by Luma and early Creators testers reveal the model’s versatility, including a man running through an Antarctic snowstorm surrounded by explosions and a ballerina performing on an ice floe in the Arctic.
Impressively, all the movements in the sample videos appear realistic and smooth – often objects move much faster and more naturally than videos from competing AI generators, which regularly seem like generated in slow motion.
The model may even create realistic versions of surreal ideas akin to browsing giraffeas user X @JeffSynthesized demonstrated. “Ray 2 is the real deal,” he says wrote on X.
Other AI video creators who have tried the recent model appear to largely agree Jerrod Lew posts on X: “There’s been improved cinematography, lighting, and realism, and it’s amazing.”
“…it’s very good!” AI video artist Heather Cooper he said.
My own tests were mixed and included more complex prompts that produced unnatural and erroneous results. But when he actually produced clips that looked more like what I had in mind in my prompts – e.g swordsmen crossing swords aboard a space station orbiting Jupiter – it was undoubtedly impressive.
Jain said Luma will even add image-to-video, video-to-video and editing capabilities to Ray2 in the future, further expanding the tool’s creative capabilities.
To have fun the launch of Ray2, Luma Labs is hosting the Ray2 Awards ceremony, offering creators the likelihood to win prizes price as much as $7,000. These include:
- Large-scale reward: The creator whose Ray2 content receives the most views on a single platform in the first week of launch will win $5,000. Applications should be submitted by January 22, 2025.
- $3,000 raffle: Creators can participate by sharing Ray2 content on social media and watching the Luma AI launch video. The deadline for participation is also January 22.
The winners of each awards can be announced on January 27. Submissions will be submitted via forms provided by Luma Labs, and creators are encouraged to make use of the hashtags #Ray2 and #DreamMachine when sharing their work.
Additionally, Luma Labs has launched an associates program that enables participants to earn commissions by promoting their tools.