Real-time interactive world model producing 720p 24fps video and audio with no fixed maximum duration, driven by commands like walking or changing weather. Auto-regressive diffusion; research preview only, via a contact form.
Runway's real-time interactive world model generating 720p/24fps video with audio, steered by text actions and camera input.
Updated EU status checked Last shown
GWM Worlds 2 at a glance
- What is GWM Worlds 2?
- GWM Worlds 2 is a model from Runway. Runway's real-time interactive world model generating 720p/24fps video with audio, steered by text actions and camera input.
- Who makes GWM Worlds 2?
- GWM Worlds 2 is made by Runway.
- Does GWM Worlds 2 work in the EU?
- Unverified. No explicit evidence found. It is a research preview with access by contact form, and no regional availability information appears on the pages provided. Last verified 14 Nov 2026.
- How much does GWM Worlds 2 cost?
- Pricing unknown. Research preview; no pricing listed for GWM Worlds 2. Access reportedly via a contact form. Runway's general plans (Standard $12/month, Pro $28/month, Max $76/month) do not list this model. Licence: Proprietary.
- Where was GWM Worlds 2 shown?
- In 1 creator video by AI Search. Most recently by AI Search on 6 Sept 2026, at 33:00 in “GPT 6 Astra, Claude Fable 5.1, Gemini 3.8, realtime Minimax, new world models: AI NEWS”.
Sources: the vendor's own pages (linked above) and creator videos at the linked timestamps. EU status is information, not legal advice; see how we check.
GWM Worlds 2 is a research-stage world model from Runway that generates interactive environments in real time: continuous 720p video at 24 fps with audio at 48,000 Hz, responding to user input as the user explores. It extends GWM Worlds, shown in December 2025, and adds generated audio and richer control over subjects and scenes.
It is built by finetuning Runway's foundational audio-video generation model on a format called WorldPrompt. WorldPrompt has a persistent layer (scene description, subjects, "laws" such as gravity or camera perspective, and a first frame) and a timestamped event stream of free-form text actions addressed to subjects or the scene, plus per-frame camera translation and rotation. The bidirectional model is then post-trained into a real-time autoregressive model, which is not limited to a fixed duration and can generate indefinitely.
Runway positions it for interactive entertainment, virtual characters, robotics and embodied-agent simulation, and generative design. It supports first-person and third-person navigation, and a player and a "director" can act at the same time. It is a research preview; the creator coverage describes access via a contact form, and the pages provided list no pricing or public API for it.
- Real-time generation of 720p video at 24 fps with 48 kHz audio
- Autoregressive model with no preset session length
- WorldPrompt input: scene, subjects, laws, first frame, timestamped text actions
- Free-form text actions addressed to subjects or the scene, with continuous camera control
- First-person and third-person navigation; player and director control simultaneously
As seen on YouTube
Watch the moment: the exact second top creators show it. Videos load only when you press play, via youtube-nocookie.com.
Similar AI tools
Closest neighbours by meaning, not just by category.
WorldPlay 2
Model
Real-time interactive world model with compressed memory and hybrid action and semantic control, built on Wan 2.2.
Solaris
Model · Runway
Runway's real-time interactive video model, built on Gen-4.5, that conditions frames on user clicks and drags.
Alaya World
Model
Interactive autoregressive world model generating minute-long playable video with real-time camera control and prompt events.
Streaming Code World Model
Model
Research code that streams video generation chunk by chunk from a rough 3D proxy, built on Code World Model.
SolarWM
Model
Open framework and dataset for turning video diffusion models into real-time, long-horizon interactive world models.
Wonder
Model · Adobe
Adobe Research video world model that builds navigable interactive worlds from an image or video at 16 FPS.