Runway, the AI startup that helped develop AI image generator Stable Diffusion, yesterday released its first mobile app giving users access to Gen-1, a video-to-video generative AI model. This app is currently only available for iOS devices.
A new app allows users to record videos from their phones and generate AI videos in minutes. You can also convert existing videos in your library with text prompts, images, or style presets.
Users can choose from a list of Runway presets such as ‘Cloudscape’ or transform videos into claymation, charcoal sketches, watercolors, origami, and more. You can also upload an image or enter your idea in the text box.
The app will generate 4 previews for the user to choose from. They select what they like best and it takes a few minutes to create the final product. Sometimes it took him two minutes to generate.
Naturally, like any AI generator, the results are not perfect and often look distorted or weird. In general, the concept of an AI video generator may seem silly. But as the technology evolves and improves over time, it can prove valuable to users.For example, content creators can spice up their social media posts.
Either way, I found Runway’s mobile app easy to use and overall fun to play with.
Here’s an example we came up with using Michael Scott’s clip from “The Office”. The text prompt I entered was “real doll”.
(Warning: the results are horrifying.)
Image credit: Runway (taken from Office clip)
I also tried “3D animation” and it worked fine.
Image credit: Runway (taken from Office clip)
Admittedly, there are some caveats beyond glitches and distorted faces.
If a user wants the free version, it has a limit of 525 credits and can only upload 5 seconds of video. Each second of the video uses 14 credits.
In the future, Runway plans to add support for longer videos, co-founder and CEO Cristobal Valenzuela told TechCrunch. The app will continue to improve and add new features, he said. added.
“We are focused on improving efficiency, quality and management. In the coming weeks and months you will see all sorts of updates, from longer outputs to higher quality video,” says Valenzuela. says.
Also note that the app doesn’t produce nudity or copyrighted work, so you can’t create videos that mimic the styles of popular IPs.
Runway’s new mobile app has two premium plans: Standard ($143.99/year) and Pro ($344.99/year). The Standard plan offers 625 credits per month and other premium features like 1080p videos, unlimited projects, and more. The Pro plan offers 2,250 credits per month and all of Runway’s 30+ AI tools.
Image credit: Runway
A month after Runway released the Gen-1 (launched in February), Runway rolled out the Gen-2 model. Perhaps a step up from Stable Diffusion, a text-to-image model, and DALL-E, Gen-2 is a text-to-video generator, allowing users to generate videos from scratch.
The runways are slowly starting to roll out access to Gen-2’s closed beta.
The app currently supports Gen-1 models, but Gen-2 will be available soon along with other AI tools in Runway, such as image-to-image generators.
Both Meta and Google have launched text-to-video generators. These are called Make-A-Video and Imagen respectively.
Since its launch in 2018, Runway has developed a range of AI-powered video editing software. The company does frame interpolation, background removal, blur effects, clean or remove audio and motion tracking, among others.
This tool has helped influencers and even film/TV studios cut down on the time they spend editing and creating videos.
For example, the visual effects team behind “Everything Everywhere All at Once” used runway technology to create scenes from the film with Evelyn (Michelle Yeoh) and Joy (Stephanie Hsu) in the multiverse, So they got to work. rock.
Additionally, the graphics team behind CBS’s “The Late Show with Stephen Colbert” used Runway to reduce editing time to just five minutes, according to Art Director Andro Buneta.
Runway also operates its entertainment and production division, Runway Studios.
