Overview
Sora is OpenAI’s text-to-video LLM capable of generating up to 60 seconds of high-definition video from text descriptions, marking a new era in AI video generation.
Key Capabilities
- Ultra-long Video: Up to 60 seconds of 1080p in a single generation
- Multi-character Interaction: Supports multiple characters interacting naturally in a scene
- World Model: Understands causal relationships in the physical world
- Video Extension: Can “animate” static images into dynamic video
Impact
Sora demonstrated the huge potential of AI to understand and generate video content, with far-reaching implications for film, advertising, and gaming.