Text-to-video generation gets most of the attention in AI video, but a lot of professional work doesn’t start from nothing — it starts from footage that already exists and needs to be changed. Runway’s Aleph 2.0, released in May alongside a new Edit Studio interface, is built specifically for that second problem: in-context editing of real clips rather than generation from a blank prompt.
Watch: Introducing Runway Aleph — A new way to edit, transform and generate video (YouTube)
How it actually works
The model takes video you already have and modifies it based on an edit made to a single frame — show it the change you want on one frame, and Aleph 2.0 propagates that change across the rest of the clip while preserving everything you didn’t ask to touch. It handles clips up to 30 seconds at 1080p, applies a single edit consistently across up to 10 cuts or scene changes without needing to be redone shot by shot, and accepts up to five keyframe anchor images for more precise control over how the edit should look. In practice that covers generating new camera angles, relighting a scene, changing weather or time of day, swapping a subject against a clean background, adding or removing objects, or transferring motion from one clip onto another.
Why editing existing footage is the harder, more useful problem
Generating a video from a text prompt is powerful, but it hands you an entirely new asset disconnected from anything you actually shot — usable for concepting, much less so for finishing real production footage. Aleph 2.0’s pitch is the opposite: it keeps your actual footage, your actual performances and framing, and changes only the specific thing you asked it to change. That’s a meaningfully closer fit to how editing and VFX work already gets done — as a series of targeted fixes and adjustments to existing material, not wholesale replacement.
The Figma Weave integration is the tell
Runway built Aleph 2.0 as a native node inside Figma Weave, Figma’s node-based AI creative workflow tool, letting reference images and frame-level decisions made in a Figma design flow directly into video edits as keyframe anchors, with no manual export step in between. That’s a small integration detail with a bigger implication: Aleph 2.0 is being positioned as a step inside an existing design-to-production pipeline, not a separate destination tool — the same distribution logic that’s made embedded AI features consistently more impactful than equally capable standalone tools this year.