Sora Video Model
World simulator and generative video model capable of rendering intricate scenes with accurate lighting, reflections, and persistent geometry.
Best for: Cinematic commercial video production, virtual world simulation, and VFX pre-visualization.
Specifications
- Provider: OpenAI
- Release date: December 2024
- Context window: Up to 20-second video clip
- Modalities: Video, Vision
- License: Proprietary Commercial API
- Pricing: Included in ChatGPT Pro ($200/mo) and API preview
- API available: Yes
Capabilities
- Spatiotemporal patch diffusion
- Complex physical simulation
- Continuous multi-camera trajectories
- High dynamic range 1080p