Models/Sora 2
provideo

Sora 2

OpenAI's long-form narrative video model — Hollywood-quality scenes.

Sora 2 is OpenAI's second-generation video model, capable of generating up to 10 seconds of photorealistic video from text. It excels at complex narrative sequences, accurate physics, and highly detailed environments. Supports 16:9 widescreen output.

Try Sora 2

Strengths

  • Long-form narrative coherence
  • Photoreal motion and lighting
  • Complex multi-character scenes
  • Accurate real-world physics
  • Expressive cinematic camera language

Sample Prompts

A woman walks through Tokyo neon-lit streets at dusk, rain reflecting neon signs

Astronaut on a lunar surface at golden hour, Earth visible in the background

A dragon soars through storm clouds over a medieval castle at night

Specs

Media typevideo
Aspect ratios16:9
Durationup to 10s
Providerfal.ai
Endpointfal-ai/sora-2/text-to-video

Related Models