"Feature","Sora (OpenAI)","Veo 3 (Google DeepMind)"
"Developer","OpenAI","Google DeepMind"
"Release Date","Announced February 2024","Announced May 2025"
"Core Technology","Diffusion model + Transformer architecture","Transformer-based, latent diffusion, multi-modal embedding"
"Input","Text prompts, images, existing video (for extension or animation)","Text prompts, images (for style/reference), video extension"
"Output","Realistic, high-fidelity videos up to 20 seconds","High-quality short videos, with advanced character consistency"
"Audio Generation","Not natively integrated (as of initial release)","Native audio generation (dialogue, soundtrack, sound effects)"
"Unique Features","Recaptioning (auto-detailed prompts), storyboard editor, remix, recut","Enhanced prompt understanding, style control, reference-based consistency"
"Availability","OpenAI API, ChatGPT Plus/Pro","Google Cloud Vertex AI, Google AI Ultra subscription"