Runway Act One

Runway Act One

Introduced with the Gen-3 Alpha of the Runway AI video generator, Act One allows users to bring characters to life from still images and reference videos, producing the most lifelike videos with human mouth movements and expressions.

Image to Video
Text to Video
Image to Video
Image

Click to upload an image

Upload JPG/PNG/WEBP images up to 10MB, with a minimum width/height of 300px.

Key Features

Animating Still Images with Videos

RunwayML Act One is designed to let you make character-focused animation clips with AI. To produce those clips, you need to combine a source video input, which gives the AI the movements and the speech or audio it needs, plus a reference image, which it uses as the baseline for the animation.

So, you can record a video of yourself reading out a script, for example, and then upload that and use an image of an animated old man as your baseline. Act One will then animate the old man image, overlay the audio from your video, and make him move and speak the same way that you (or your subject) did in the input video.

Learn more: How to use Runway's Act One?

Lifelike Movements & Expressions

One of the best aspects of Runway’s Act One is how lifelike and realistic the output is. It has extraordinary levels of understanding facial anatomy and movement, and it uses the movement seen in the input video and copies that over to the final animation. It’s almost like doing motion capture, but without any expensive equipment or time-consuming processes.

A Range of References

When it comes to picking the base or foundation of your animation, Runway gives you a lot of options. It actually offers dozens of different reference images and short video clips in a range of styles, like 3D animated, 2D animated, illustrated, and photorealistic, and these reference characters aren't just limited to human ones.

You can also make animations involving animal faces, robots, and more, so you can effectively create realistic animations that look however you want, from Disney-style princesses to lifelike people. The tool also allows you to upload your own image, should you prefer, opening even more opportunities for creativity and expression.

Motion Intensity Options

Before generating your animation with Act One, you also get to configure the Motion Intensity setting, which ranges from 1 (the least amount of motion) to 5 (the most). This allows you to set exactly how much movement you get in your final video, so you can make them more energetic and dynamic or more relaxed and calm, according to the desired mood.

Motion Intensity value Output
1
motion-1
3
motion-2
5
motion-3

Video reviews

Other Reviews on X

FAQs

What is Act-One Runway?

Act-One is an innovative AI tool of Runway for generating expressive character performances in videos. Launched in October 2024 as part of Runway's Gen-3 Alpha platform, Act-One allows users to create animated characters based on simple video input.

What can you do with Runway's Act One?

You can use Runway Act One to make AI character animations, using video input and image references. It works a lot like motion capture, converting the movements from a video onto an AI animation. You can use this for narrative videos, tutorials, advertisements, and more.

What are some best practices for using Act One?

Act One relies on good quality video input to produce the best quality videos. So you need to make sure the videos you upload are good quality, with a single subject facing forward and clearly visible. Avoid excessive head and body movements, and use clear and pronounced facial movements and expressions.

Is Runway Act-One free?

To be precise, Runway takes 10 credits for every second on the Gen-3 Alpha model or five per second if you're using Gen-3 Alpha Turbo. The longer your clips, the more credits you’ll use.

What is the maximum length of an Act One video?

The maximum length of Act One videos is 30 seconds. Minimum length is five seconds.

What to do if Act One doesn't work?

There are a few troubleshooting steps you can try if Act One isn't giving the right results. The best way to usually fix it is to record and upload a better quality video input with good lighting, a clearly defined subject, and minimal body and background movement.

Experience the Ultimate AI Video Generator!

Experience the Ultimate AI Video Generator!

With Pollo AI, you can easily access top AI video models like Kling AI, Runway, Pixverse AI, Vidu AI, Luma AI, and more—all in one place, for free.