AI GlossaryGWords you meet while using AI
GWM-1
The earlier version of Runway's world model that renders a walkable virtual space in real time.
In plain words
GWM-1 was the earlier version of the world model built by Runway. It works by continuously redrawing the scene in front of you based on your camera position and inputs. Rather than producing a finished video clip in advance, it generates the next frame on the fly as you move through the space.
At this stage, though, it was visuals only, with no sound. There was no background audio or footstep sound, just a visual experience — a kind of silent-movie version of a virtual space.
The successor model that came later added synchronized sound and the ability to keep exploring continuously. GWM-1 is best understood as the stage before that.
How it shows up in the news
The article introduces the successor, GWM Worlds 2, and refers to the earlier version not by name but as "the previous version." The line "the biggest change from the previous version is this sound" is referring to GWM-1. It's easy to mistake GWM-1 for a finished game or a video production tool, but it's actually closer to a research model that generates a walkable virtual space in real time.
See also
Stories using this term
- Runway unveils playable world model with soundCreative · 2026.09.05
- LTX-2.5, open video generation model runs on a single consumer RTX GPUAI · 2026.08.12
- Runway unveils Solaris, which redraws the screen with every clickAI · 2026.09.01
- Alibaba unveils Wan-Animate-2, adds real-time streaming generationAI · 2026.08.11
- World Labs' Atlas Generates and Reconstructs 3D Worlds From a Few PhotosAI · 2026.09.03
- Runway announces winners of second 'Ads for Products That Don't Exist' contestCreative · 2026.08.15
