New York-based artificial intelligence company Runway has launched Aleph, a video-to-video AI model designed specifically for post-production workflows. Announced on July 25, the tool enables users to modify existing video footage through simple text prompts, marking a shift away from traditional generation-based AI video tools.
Runway Aleph is a state-of-the-art in-context video model, setting a new frontier for multi-task visual generation, with the ability to perform a wide range of edits on an input video such as adding, removing, and transforming objects, generating any angle of a scene, and modifying style and lighting, among many other tasks.
The model's core capabilities include generating new camera angles from existing footage, object manipulation, and environment transformation. The tool can also generate additional frames beyond the original video and create alternative perspectives, such as extreme close-ups or wide-angle shots. All of these transformations are performed while preserving the original context and continuity of the footage.
Technical features include precise object removal and addition, with the AI automatically adjusting lighting, shadows, and perspective to maintain realism. The model can transform lighting conditions — changing harsh noon light to golden hour or converting day scenes to night — while updating colour temperature and reflections across the entire frame.
One of Aleph's standout features is the ability to generate entirely new camera angles for a video. With prompts like "Generate a medium full shot of the subject," Aleph creates fresh perspectives, giving editors what Runway calls "endless coverage." The tool also includes a green screen function that can isolate subjects with precise edge detection, preserving fine details for compositing work.
Aleph enters a competitive market dominated by tech giants. Adobe has focused on integrating its "commercially safe" Firefly AI directly into Premiere Pro, a strategy designed to appeal to enterprises wary of copyright disputes. Meanwhile, models like OpenAI's Sora and Google's Veo 2 have set a high bar for text-to-video realism, creating entire scenes from prompts.
Early access will begin rolling out shortly for all our Enterprise and Creative Partners. With broader access coming soon to all users. The company has not disclosed technical specifications such as supported video lengths, aspect ratios, or pricing structures.
Runway's move comes amid growing industry adoption of AI tools, with the company securing partnerships with Hollywood studio Lionsgate and IMAX for AI-generated film screenings.