Runway will release a new AI (AI) video generation model on Friday, allowing you to edit elements from other input video. A video-to-video AI model called Aleph (the first letter of the Hebrew alphabet) can perform a wide range of editing on input, including adding, removing, and converting objects. Adding new angles and next frames. Changes in environment, seasons and times. Transforms styles. The New York-based AI company said the video model will soon be rolled out to its companies and creative customers as well as users of the platform.
Runway's AlephAI model can edit videos
AI video generation models have come a long way. We have moved from generating animated scenes for a few seconds to generating full-scale video for stories and even audio. Runway is at the forefront of this innovation using video generation tools, and is now used by producers such as Netflix, Amazon and Walt Disney.
Currently, AI companies are releasing a new model called Aleph, which allows you to edit input videos. This is a video-to-video model that can be manipulated and generated by manipulating a wide range of elements. In a post on X (formerly known as Twitter), the company called it the state-of-the-art (SOTA) internal video model that allows you to convert input video with a simple description prompt.
Introducing Runway Aleph, a new way to edit, convert and generate videos.
Aleph is a state-of-the-art in-context video model with the ability to set up new frontiers for multitasking visual generation and perform a wide range of editing on input videos, such as adding and removing. pic.twitter.com/zgdwyedmqm
– Runway (@runwayml) July 25th, 2025
In a blog post, AI companies showed some of the features they offer when ALEPH becomes available. Runway says the model will first be offered to corporate and creative customers, and will then be widely released to all users in a few weeks. However, phrasing does not make it clear whether free tier users will have access to the model or whether they will be booked for paid subscribers.
Aleph can shoot input video and generate new angles and views of the same scene, all the way to its functionality. Users can request reverse low shots, extreme close-ups from the right side, or wide shots of the entire stage. You can also use the input video as a reference to generate the next frame of the video based on the prompt.
One of the most impressive capabilities of AI video models is to convert environments, locations, seasons and times. This means that users can upload park videos on sunny mornings, allowing AI models to add rain, sandstorms and snowfall effects, or make them look like nighttime. These changes are made while keeping all the other elements of the video intact.
Aleph can add objects to the video, remove reflections, buildings, etc., completely change objects and materials, change the appearance of characters, and re-order objects. Additionally, Runway claims that AI models can perform certain movements in the video (think flying drone sequences) and apply them to different settings.
Currently, Runway does not share technical details about the AI model, including the supported length of input video, supported aspect ratios, and application programming interface (API) pricing. These may be shared when the company officially releases the model.
