Gen-3 Alpha was designed from the ground up for creative applications, which enables it to understand and generate a wide range of styles and artistic instructions. pic.twitter.com/QoACOW043Q
— Cristóbal Valenzuela (@c_valenzuelab) June 17, 2024
Runway, a New York City-based AI startup, has unveiled its latest breakthrough in generative AI technology, the Gen-3 Alpha model. This new model can produce hyper-realistic video clips up to 10 seconds long, with high precision in emotional expressions and camera movements. Gen-3 Alpha is a significant advancement in video generation.
Runway just announced their new AI video generation model called Gen-3 Alpha.
Look at these 10 insane examples, all AI 🤯
1. Bald man is happy to have a new wig pic.twitter.com/ftZAMxlhay
— Alvaro Cintas (@dr_cintas) June 17, 2024
Runway describes it as the “first of an upcoming series of models” designed on a new infrastructure for large-scale multimodal training. This development is a step towards building “General World Models” that can represent and simulate a wide range of real-world situations and interactions. A Runway spokesperson explained the technical prowess of the model: “This initial rollout will support 5 and 10-second generations, with noticeably faster generation times.
A 5-second clip takes 45 seconds to generate, and a 10-second clip takes 90 seconds to generate.”
This is wild.
Runway AI just dropped Gen-3 Alpha that generates AI video just from text and images.
A step toward building General World Models.
100% AI 🤯
1. Prompt: A middle-aged sad bald man becomes happy as a wig of curly hair and sunglasses fall suddenly on his head. pic.twitter.com/bWUL31pZ0e
— Min Choi (@minchoi) June 17, 2024
Beastmode.
Test ideas quickly by generating 5-second video clips in ~45 seconds.
Generated with @runwayml GEN-3 Alpha. pic.twitter.com/5WaDrUrlJT
— Nicolas Neubert (@iamneubert) June 18, 2024
Demo videos are available on Runway’s website and social media, although the precise release date for Gen-3 Alpha has not been disclosed.
Runway launches latest AI model
The model will first be accessible to paying subscribers, with plans to offer it to free-tier users later.
Subscription pricing starts at $15 per month or $144 per year. Anastasis Germanidis, Runway’s co-founder, stated on social media that Gen-3 Alpha “will soon be available in the Runway product and will power all existing modes (text-to-video, image-to-video, video-to-video) along with new capabilities enabled by the more advanced model.”
Since the launch of Gen-2 in 2023, Runway has learned that video diffusion models significantly benefit from scaling, leading to powerful visual world representations. Gen-3 Alpha follows this trend, boasting joint training on videos and images through a collaborative effort involving research scientists, engineers, and artists.
However, the specific data sets used for training remain undisclosed. Runway has already partnered with prominent entertainment and media organizations to create custom versions of Gen-3, tailored to meet specific artistic and narrative needs. Previous Runway collaborations include notable projects like the films “Everything, Everywhere, All at Once” and “The People’s Joker.”
With the introduction of Gen-3 Alpha, Runway aims to maintain its position as a leading player in the rapidly advancing generative AI video creation space.