RunwaySpecialist modelPreview
GWM Worlds 2
n/ano search or usage data
Summary
GWM Worlds 2 is Runway’s research preview for interactive worlds. The autoregressive diffusion model generates continuous 720p video at 24 frames per second with synchronized 48,000 Hz audio. Text actions change the scene, subjects, and camera motion during the simulation.
Runway describes a continuous world without a fixed clip duration. Its video and audio input claims are contradictory. One demonstration continues an eight-second video with audio, while the limitations section rules out those inputs. Image references beyond the first frame are unsupported. General API access is not confirmed.
On this page
Similar models
More models of the same kind as GWM Worlds 2, chosen by model line, developer, size, price and release date.
- Praxis-1Runway · same developer
- Gemini 3.8 Flash CyberGoogle DeepMind · released around the same time
- GPT-Rosalind-5.5OpenAI · released around the same time
- MAI-Cyber-1-FlashMicrosoft AI · released around the same time
- Perceptron Mk1.5Perceptron · released around the same time
- GPT-5.6 CyberOpenAI · released around the same time
- EmbeddingGemma 2Google DeepMind · released around the same time
- Unlimited OCR 2Baidu · released around the same time
All specifications
Architecture
Watermarking and provenance
AI watermarking refers to invisible signals embedded in the output. Visible marks and provenance metadata are listed separately. Unknown means there is insufficient evidence for that particular field.
- AI watermarking
- Unknown
- no source
- Watermarking technique
- Not known
- no source
Licence and access
- Shutdown
- None announced
- no source
Further details
- Interaction
- Text actions control the scene, subjects, and camera motion in real time
- Source: Interaction
- Limitations
- No image references beyond the first frame; imperfect long-term memory and detail degradation with fast camera motion
- Source: Limitations
- Video and audio inputs
- Provider claims are contradictory. Runway demonstrates continuing an eight-second video with audio, but lists prefilled video and audio as unsupported in its limitations. The exact scope remains unresolved.
- Source: Video and audio inputs
Sources and data date
| Type | Evidence and data date |
|---|---|
| Research status | Research date October 4, 2026. 2 source URLs checked. This documents the inspected sources, not an exhaustive inventory of every publication. |
| Not published | quality-benchmark Original wording The 720p, 24 fps, 48 kHz, and eight-second demonstrations are output specifications, not quality benchmarks. |
| Not found in the inspected sources | exact-model-quality-benchmark Note Checked official introduction contains no standardized GWM Worlds 2 benchmark score; demonstrations and product-rate claims are not quality scores. |
| Not found in the inspected sources | gwm-worlds-2-quality-score Note No exact quantitative quality result was found. Published 720p, 24 fps, audio sample rate and clip duration describe output specifications, not measured comparative performance. |
| Not published | Published quantitative robotics/world-model benchmark Note The current official announcement gives output specifications and qualitative action-control claims but no numerical benchmark metric, dataset result, or sample count. Resolution/frame rate are not benchmark scores. |
| Additional source | Runway GWM Worlds 2 announcement (retrieved October 3, 2026; October 4, 2026; October 15, 2026; October 16, 2026; October 28, 2026) · Runway GWM Worlds 2 announcement · Editorial description reviewed October 3, 2026 |